tag:theconversation.com,2011:/global/topics/big-data-3963/articlesBig data – The Conversation2024-02-08T03:40:45Ztag:theconversation.com,2011:article/2228832024-02-08T03:40:45Z2024-02-08T03:40:45ZSolving the supermarket: why Coles just hired US defence contractor Palantir<p>What does the Australian supermarket chain Coles have in common with the CIA? As of last week, both are clients of <a href="https://www.palantir.com/about/">Palantir Technologies</a>, a US tech company “focused on creating the world’s best user experience for working with data”.</p>
<p>In a three-year deal, Coles plans to deploy Palantir’s tools across more than 840 supermarkets to <a href="https://www.afr.com/world/north-america/coles-brings-in-pentagon-s-palantir-for-cost-cutting-20240202-p5f1tq">cut costs</a> and “redefine how we think about our workforce”. </p>
<p>The tech company, named after magical seeing stones from the Lord of the Rings, offers comprehensive software that collects, organises and visualises a client’s data in “<a href="https://hcommons.org/deposits/item/hc:34683/">one platform to rule them all</a>”. For an intelligence agency, Palantir’s tools might help <a href="https://www.youtube.com/watch?v=QIT4hv4tnek">identify a terror cell</a> through phone calls and financial transactions; in a healthcare organisation, they might find ways to save money by <a href="https://www.youtube.com/watch?v=tljWVIUbulg">shortening emergency department stays</a>.</p>
<p>For Coles, the <a href="https://www.businesswire.com/news/home/20240201306532/en/Palantir-Partners-with-One-of-Australia%E2%80%99s-Leading-Retailers">goal</a> is to “optimise its workforce” by analysing “over 10 billion rows of data, comprising each store, team member, shift and allocation across all intervals in a day, every day”. </p>
<p>The announcement is <a href="https://www.itnews.com.au/news/coles-to-run-palantir-analytics-suite-across-its-supermarkets-604698">linked</a> to Coles’ plan to save a billion dollars over the next four years, and follows a 2019 <a href="https://news.microsoft.com/en-au/features/coles-takes-trip-down-the-aisles-with-microsoft/">big data deal with Microsoft</a>, an effort to build <a href="https://www.afr.com/companies/retail/coles-faces-ocado-delivery-and-cost-blowout-20230818-p5dxik">robotic delivery centres</a>, and the introduction of <a href="https://www.news.com.au/finance/business/retail/experts-warn-about-customer-privacy-after-drastic-security-moves-by-supermarkets/news-story/66e0ee85491eacf49fe18e30ee49197e">customer-tracking cameras</a> and other high-tech security measures.</p>
<h2>The Palantir process</h2>
<p>What might this Palantir–Coles collaboration look like in practice? </p>
<p>Typically, Palantir first sends out “forward-deployed engineers” to begin work with an organisation’s data, which is often messy, incomplete and fragmented. These engineers work with different branches and stakeholders to bring the data together into a single compatible whole called “<a href="https://www.palantir.com/explore/platforms/foundry/ontology/">The Ontology</a>”, which contains all the information deemed relevant. </p>
<p>Then the data can be fed into Palantir’s platforms – in this case, customisable software called <a href="https://www.palantir.com/platforms/foundry/">Foundry</a> and the <a href="https://www.palantir.com/platforms/aip/">Artificial Intelligence Platform</a>.</p>
<hr>
<p>
<em>
<strong>
Read more:
<a href="https://theconversation.com/how-tech-billionaires-visions-of-human-nature-shape-our-world-144016">How tech billionaires' visions of human nature shape our world</a>
</strong>
</em>
</p>
<hr>
<p>The platforms let clients explore the data through <a href="https://betterprogramming.pub/palantir-foundry-the-data-operating-system-that-is-not-talked-about-enough-9fb1c98a6b3d">dense but user-friendly interfaces</a> populated by columns and rows, boxes and lines. The Artificial Intelligence Platform also brings ChatGPT-like language models into the mix. </p>
<p>Users might compare earnings between branches, flag a store that seems inefficient, or identify an upcoming period of high spending based on historic patterns. </p>
<p>All of this probably seems banal, or even boring. It’s certainly less overtly problematic than Palantir’s work with governments and law enforcement, which has been slammed for enabling <a href="https://slate.com/technology/2020/09/palantir-ice-deportation-immigrant-surveillance-big-data.html">data-driven deportation</a> or <a href="https://theintercept.com/2021/01/30/lapd-palantir-data-driven-policing/">racist policing</a>, and seen the company described as “<a href="https://slate.com/technology/2020/01/evil-list-tech-companies-dangerous-amazon-facebook-google-palantir.html">evil</a>”. </p>
<hr>
<p>
<em>
<strong>
Read more:
<a href="https://theconversation.com/high-tech-surveillance-amplifies-police-bias-and-overreach-140225">High-tech surveillance amplifies police bias and overreach</a>
</strong>
</em>
</p>
<hr>
<p>However, the deal doesn’t need to be overtly malevolent to be meaningful. A technology of surveillance and control is quietly <a href="https://www.routledge.com/Countering-the-Cloud-Thinking-With-and-Against-Data-Infrastructures/Munn/p/book/9781032374154">becoming infrastructure</a>, moving from front-page news to something ticking along silently in the background. In this sense, Palantir shifts from the visible to the operational, imperceptibly but powerfully shaping the lives and livelihoods of Australian supermarket employees and shoppers. </p>
<h2>Optimising the workforce</h2>
<p>We can briefly sketch out three implications of the deal.</p>
<p>First, by inking this deal, Coles frames itself as future-forward and logistically driven. Groceries and grocery-store labour become more data, just like the hedge funds, healthcare, or immigrants that other Palantir clients coordinate. </p>
<hr>
<p>
<em>
<strong>
Read more:
<a href="https://theconversation.com/coles-and-woolworths-are-moving-to-robot-warehouses-and-on-demand-labour-as-home-deliveries-soar-166556">Coles and Woolworths are moving to robot warehouses and on-demand labour as home deliveries soar</a>
</strong>
</em>
</p>
<hr>
<p>Supermarkets have been under fire over the past year for <a href="https://www.theguardian.com/business/2023/may/22/australias-big-supermarkets-increased-profit-margins-through-pandemic-and-cost-of-living-crisis-analysis-reveals">increasing profit margins</a> through a pandemic and cost-of-living crisis, and accused of <a href="https://www.theguardian.com/business/2023/jun/05/coles-woolworths-court-accused-of-underpaying-workers">underpaying workers</a>. </p>
<p>The Palantir deal continues this extractive trajectory. Rather than paying workers more or passing savings onto customers, Coles has chosen to invest millions in technology that will “address workforce-related spend” as part of a <a href="https://theshout.com.au/national-liquor-news/coles-ceo-outlines-strategies-for-christmas-and-beyond/">larger effort to cut costs</a> by a billion dollars over the next four years. Food (and the labour needed to grow, pack and ship it) is transformed from a human need to an optimisation problem. </p>
<h2>A walled garden</h2>
<p>Second, dependence. As <a href="https://meson.press/books/ferocious-logics/">my own research found</a>, Palantir clients tend to enjoy the all-encompassing data and new features but also become dependent on them. Data mounts up; new servers are needed; licensing fees are high but must be paid. </p>
<p>Much like Apple or Amazon, Palantir’s services excel at creating “vendor lock-in”, a perfect walled garden which clients find hard to leave. This pattern suggests that, over the next three years, Coles will increasingly depend on Silicon Valley technology to understand and manage its own business. A company that sells a quarter of Australia’s groceries may become operationally reliant on a US tech titan.</p>
<h2>A way of seeing</h2>
<p>Finally, vision. What Palantir sells is fundamentally a way of seeing. Its dashboards promise <a href="https://meson.press/books/ferocious-logics/">a God’s eye view</a> that can stretch across an entire organisation or zoom in to granular detail to locate that “needle in the haystack” insight. </p>
<p>The claim is that this data-driven view is a shortcut to <a href="https://www.tandfonline.com/doi/abs/10.1080/1472586X.2014.887268">total knowledge</a>, a way to map every operation, reveal every important element, and identify every inefficiency. </p>
<figure class="align-center zoomable">
<a href="https://images.theconversation.com/files/574260/original/file-20240208-23-vimnk3.jpeg?ixlib=rb-1.1.0&q=45&auto=format&w=1000&fit=clip"><img alt="A complex diagram illustrating the Palantir 'ontology' and how it can be used in an organisation." src="https://images.theconversation.com/files/574260/original/file-20240208-23-vimnk3.jpeg?ixlib=rb-1.1.0&q=45&auto=format&w=754&fit=clip" srcset="https://images.theconversation.com/files/574260/original/file-20240208-23-vimnk3.jpeg?ixlib=rb-1.1.0&q=45&auto=format&w=600&h=398&fit=crop&dpr=1 600w, https://images.theconversation.com/files/574260/original/file-20240208-23-vimnk3.jpeg?ixlib=rb-1.1.0&q=30&auto=format&w=600&h=398&fit=crop&dpr=2 1200w, https://images.theconversation.com/files/574260/original/file-20240208-23-vimnk3.jpeg?ixlib=rb-1.1.0&q=15&auto=format&w=600&h=398&fit=crop&dpr=3 1800w, https://images.theconversation.com/files/574260/original/file-20240208-23-vimnk3.jpeg?ixlib=rb-1.1.0&q=45&auto=format&w=754&h=500&fit=crop&dpr=1 754w, https://images.theconversation.com/files/574260/original/file-20240208-23-vimnk3.jpeg?ixlib=rb-1.1.0&q=30&auto=format&w=754&h=500&fit=crop&dpr=2 1508w, https://images.theconversation.com/files/574260/original/file-20240208-23-vimnk3.jpeg?ixlib=rb-1.1.0&q=15&auto=format&w=754&h=500&fit=crop&dpr=3 2262w" sizes="(min-width: 1466px) 754px, (max-width: 599px) 100vw, (min-width: 600px) 600px, 237px"></a>
<figcaption>
<span class="caption">Palantir promises a ‘total view’ of an organisation that allows full control and optimal decision-making.</span>
<span class="attribution"><a class="source" href="https://blog.palantir.com/connecting-ai-to-decisions-with-the-palantir-ontology-c73f7b0a1a72">Palantir</a></span>
</figcaption>
</figure>
<p>Yet the data inevitably excludes significant social, financial and environmental information. The sweat of workers struggling to pack at pace, the belt-tightening of consumers struggling to make ends meet, and the struggle of farmers to survive unexpected climate impacts will go untracked. </p>
<p>Such details never appear on the platform – and if they’re not data, they don’t matter. Will Palantir’s data-driven myopia translate to how Coles views its workers and customers? </p>
<p>By placing Palantir at the heart of its operations, Coles quietly smuggles in several key assumptions: that food is a commodity to be optimised, that paying for labor is a risk rather than a responsibility, and that data can capture everything of importance. At a time of <a href="https://www.aph.gov.au/Parliamentary_Business/Committees/House/Agriculture/FoodsecurityinAustrali/Report/Chapter_7_-_Food_insecurity">increased food insecurity</a>, Australians should strongly question whether this is the direction one of our major grocery providers should take.</p><img src="https://counter.theconversation.com/content/222883/count.gif" alt="The Conversation" width="1" height="1" />
<p class="fine-print"><em><span>Luke Munn does not work for, consult, own shares in or receive funding from any company or organisation that would benefit from this article, and has disclosed no relevant affiliations beyond their academic appointment.</span></em></p>Coles plans to ‘optimise its workforce’ with big data and AI tools from a controversial tech company.Luke Munn, Research Fellow, Digital Cultures & Societies, The University of QueenslandLicensed as Creative Commons – attribution, no derivatives.tag:theconversation.com,2011:article/2186292023-11-30T10:21:06Z2023-11-30T10:21:06ZPalantir: privacy fears over handing NHS data to US defence provider show how lack of trust is holding back much-needed reform<figure><img src="https://images.theconversation.com/files/562145/original/file-20231128-29-yf36gf.jpg?ixlib=rb-1.1.0&rect=613%2C49%2C6769%2C4978&q=45&auto=format&w=496&fit=clip" /><figcaption><span class="caption">
</span> <span class="attribution"><a class="source" href="https://www.shutterstock.com/image-photo/clerk-leafing-through-stored-folders-looking-2254682601">Shutterstock/Marian Weyo</a></span></figcaption></figure><p>Controversial US tech company Palantir <a href="https://www.england.nhs.uk/2023/11/new-nhs-software-to-improve-care-for-millions-of-patients/">has been awarded a £330 million contract</a> to create a new system for sharing data – including patients’ medical details – within the NHS in England. </p>
<p>The move has been <a href="https://www.england.nhs.uk/2023/11/new-nhs-software-to-improve-care-for-millions-of-patients">welcomed by NHS leaders</a> as a way to enable healthcare workers to access live healthcare data at the “touch of a button”. But doctors’ organisations and human rights charities have <a href="https://www.theguardian.com/society/2023/nov/21/patient-privacy-fears-us-spy-tech-firm-palantir-wins-nhs-contract">expressed concerns</a> <a href="https://www.nytimes.com/2023/11/21/business/palantir-nhs-uk-health-contract-thiel.html">about the contract</a> and Palantir, including whether patient data would be suitably protected.</p>
<p>The NHS is in desperate need of a better way to share information between the many care organisations that comprise it. The inability of its many existing data systems to talk to each other can lead to delays in care, poor understanding of local health service needs and hide inequalities in <a href="https://www.health.org.uk/publications/long-reads/how-better-use-of-data-can-help-address-key-challenges-facing-the-nhs">who gets care</a>.</p>
<p>However, our research shows the UK public are currently ambivalent about their medical data being handled by private companies. So actions that further increase mistrust risk holding back these vital reforms.</p>
<p>Data on the health of UK citizens is held in many databases across GP practices, hospitals, health authorities, care homes, pharmacies and many other organisations. This means a doctor seeing a patient from a different area of the UK would typically not be able to rapidly access their hospital records.</p>
<p>But we know things can be different. Close to real-time <a href="https://www.thelancet.com/journals/landig/article/PIIS2589-7500(22)00147-9/fulltext#seccestitle30">information</a> on the trajectory of the COVID pandemic helped coordinate a <a href="https://www.instituteforgovernment.org.uk/publication/government-data-sharing-pandemic">nationwide response</a>. (And some of these new capabilities were actually <a href="https://www.england.nhs.uk/contact-us/privacy-notice/how-we-use-your-information/covid-19-response/nhs-covid-19-data-store/">facilitated by Palantir</a>.)</p>
<p>Sadly, the post-pandemic benefits of data sharing for patients with other conditions have been short-lived. Regulations have, for the most part, returned to pre-pandemic restrictions and data put back in silos.</p>
<h2>Palantir’s trust problem</h2>
<p>So a new system for sharing data across the NHS is vital. The issue highlighted by awarding the new contract to Palantir is how important public trust is to successful reform, and how missteps could damage that trust.</p>
<p>Palantir’s detractors, such as <a href="https://www.youtube.com/watch?v=HlHPOH2s4Xg">Conservative MP David Davis</a>, say questions remain over whether it can be trusted with private information due, in part, to its <a href="https://www.washingtonpost.com/business/2019/08/22/war-inside-palantir-data-mining-firms-ties-ice-under-attack-by-employees/">record</a> of working with intelligence, immigration and military organisations in the US and its founder’s financial backing for the successful 2016 Trump <a href="https://www.washingtonpost.com/business/2019/08/22/war-inside-palantir-data-mining-firms-ties-ice-under-attack-by-employees/">presidential campaign</a>. </p>
<p>Privacy concerns have also been raised, by groups <a href="https://www.theguardian.com/society/2023/nov/20/nhs-england-gives-key-role-in-handling-patient-data-to-us-spy-tech-firm-palantir">including the British Medical Association</a>, about whether confidential data will be seen by Palantir and other organisations outside of the NHS. </p>
<p>In a <a href="https://www.youtube.com/watch?v=HlHPOH2s4Xg">television interview</a> with Palantir’s CEO Alex Karp shortly before the announcement, concerns were raised about them selling NHS patient data. In response, Karp <a href="https://www.bbc.co.uk/news/technology-67254010">told the BBC</a>: “We’re the only company of our size and scale that doesn’t buy your data, doesn’t sell your data, doesn’t transfer it to any other company … That data belongs to the government of the United Kingdom.”</p>
<figure>
<iframe width="440" height="260" src="https://www.youtube.com/embed/HlHPOH2s4Xg?wmode=transparent&start=0" frameborder="0" allowfullscreen=""></iframe>
</figure>
<p>However, if the public start to feel that their data will not be protected, it could erode the already limited trust they have in private companies to be involved in the NHS.</p>
<p>Our research is attempting to identify public attitudes to NHS data sharing. We work on Datamind – the Medical Research Council and Health Data Research UK’s Mental Health Data Hub. We have published <a href="https://bmjopen.bmj.com/content/12/4/e057579">a large-scale assessment</a> of UK public opinion on NHS data sharing, which consulted almost 30,000 people. While the public are highly supportive of data sharing with the NHS, charities and university researchers, they are less trustful of private companies. </p>
<p>Meanwhile, previous examples of NHS-industry data sharing collaborations have not helped and show what can happen if trust isn’t secured. <a href="https://www.england.nhs.uk/wp-content/uploads/2015/07/care-data-Quick-reference-guide-v1.0.pdf">Care.data</a> was a programme that aimed to increase the range of health information collected across all NHS funded services, including general practice, for service planning and research. </p>
<p>It was explicitly stated that information would only be shared with industry for the benefit of health and care, such as developing new drugs, but not where it was solely for commercial purposes, such as insurance or marketing. Despite this, the <a href="https://www.theguardian.com/technology/2016/jul/06/nhs-to-scrap-single-database-of-patients-medical-details">programme was scrapped</a> because of remaining concerns over who would be allowed to see confidential patient data, what it would be used for and whether patients would have the ability to opt-out. </p>
<h2>What can be done to improve trust?</h2>
<p>We believe that the sharing of patient data across NHS datasets could have enormous public health benefits that improve outcomes across the whole of medicine. But to achieve that, there has to be trust in the private companies that will inevitably be involved.</p>
<p>As we sit in our clinics, medical staff view and enter electronic health records on a Dell computer running Windows software manufactured by Microsoft that uses a patient record management system provided by the <a href="https://www.intersystems.com/uk/about-us/">InterSystems</a> Corporation. In other words, many services in the NHS are already provided by commercial companies. </p>
<p>There is a pressing need for the public to have the knowledge, language and skills to properly engage in nuanced conversations about the use of their data and to build trust in its responsible use where appropriate. </p>
<p>We developed an online <a href="https://datamind.org.uk/patients-and-public/data-literacy-short-course-2/">data literacy course</a> for the public with our <a href="https://datamind.org.uk/patients-and-public/the-super-research-advisory-group/">Research Advisory Group</a> and the patient engagement charity <a href="https://mcpin.org/">McPin</a>. We also developed a <a href="https://datamind.org.uk/glossary/">glossary tool</a> so people could quickly look up data science terms in easily understandable language. These allow people to effectively understand the information presented to them and ask difficult questions. </p>
<p>These conversations need to happen now. Otherwise we will lurch from one apparently misinformed data sharing crisis to another, further eroding public trust. This risks all the opportunities responsible safe and secure data sharing could provide. </p>
<p>Fundamentally, however, both government and industry need to prove that they are worthy of public trust. They need to commit to engaging the public as equal partners on the potential beneficiaries of this and other data sharing initiatives. We all need to know who stands to benefit from sharing our health data so that we can make truly informed decisions.</p><img src="https://counter.theconversation.com/content/218629/count.gif" alt="The Conversation" width="1" height="1" />
<p class="fine-print"><em><span>Andrew M McIntosh receives funding from The Wellcome Trust, UKRI, The European Commission and the US National Institutes of Health. He is the Chief Scientist of Datamind, the Health Data Research UK Hub for Mental Health. </span></em></p><p class="fine-print"><em><span>Ann John receives funding from the MRC, MQ and Wellcome and is Principal investigator and Co-director of Datamind. She is a former Trustee at Samaritans and MQ.</span></em></p>The public are ambivalent about their medical data going to private companies, and missteps could erode vital trust.Andrew M McIntosh, Professor of Biological Psychiatry, The University of EdinburghAnn John, Clinical Professor of Public Health and Psychiatry, Swansea UniversityLicensed as Creative Commons – attribution, no derivatives.tag:theconversation.com,2011:article/2148192023-11-22T13:17:53Z2023-11-22T13:17:53ZDigitized records from wildlife centers show the most common ways that humans harm wild animals<p>At hundreds of wildlife rehabilitation centers across the U.S., people can learn about wild animals and birds at close range. These sites, which may be run by nonprofits or universities, often feature engaging exhibits, including “ambassador” animals that can’t be released – an owl with a damaged wing, for example, or a fox that was found as a kit and became accustomed to being fed by humans. </p>
<p>What’s less visible are the patients – sick and injured wild animals that have been admitted for treatment.</p>
<p>Each year, people bring hundreds of thousands of sick and injured wild animals to wildlife rehab centers. Someone may find an injured squirrel on the side of the road or notice a robin in their backyard that can’t fly, and then call the center to pick up an animal in distress.</p>
<p>We study <a href="https://www.researchgate.net/profile/Tara-Miller-8">ecology</a> and <a href="https://scholar.google.com/citations?user=XfgB_BUAAAAJ&hl=en">biology</a>, and recently used newly digitized records from wildlife rehabilitation centers to identify the human activities that are most harmful to wildlife. In the largest study of its kind, we reviewed 674,320 records, mostly from 2011 to 2019, from 94 centers to paint a comprehensive picture of threats affecting over 1,000 species across much of the U.S. and Canada. </p>
<p>Our findings, <a href="https://doi.org/10.1016/j.biocon.2023.110295">published in the journal Biological Conservation</a>, point to some strategies for reducing harm to wildlife, especially injuries caused by cars.</p>
<figure>
<iframe width="440" height="260" src="https://www.youtube.com/embed/SEVqsMsvQws?wmode=transparent&start=0" frameborder="0" allowfullscreen=""></iframe>
<figcaption><span class="caption">The Wildlife Rehabilitation Center of Minnesota, the largest independent rehab center in the U.S., treats over 1,000 sick and injured animals yearly.</span></figcaption>
</figure>
<h2>Tracking the toll</h2>
<p>Humans are responsible for the deaths and injuries of billions of animals every year. Bats and birds fly into buildings, power lines and wind turbines. Domestic cats and dogs kill backyard birds and animals. Development, farming and industry alter or destroy wild animals’ habitats and expose wildlife to toxic substances like lead and pesticides. Extreme weather events linked to climate change, such as flooding and wildfires, can be devastating for wildlife.</p>
<p>Most Americans support <a href="https://www.ifaw.org/press-releases/survey-majority-americans-support-candidate-values-protection-endangered-species">protecting threatened and endangered species</a>, and <a href="https://biologicaldiversity.org/w/news/press-releases/survey-most-americans-believe-human-population-driving-wildlife-extinctions-2020-11-12/">recognize that human activities can harm wildlife</a>. But it is surprisingly difficult to determine which activities are most harmful to wildlife and identify effective solutions. </p>
<p>Information from wildlife rehab centers across the U.S. can help fill in that picture. When an animal is brought into one of these centers, a rehabilitator assesses its condition, documents the cause of injury or illness if it can be determined, and then prepares a treatment plan. </p>
<p>Wildlife rehabbers may be veterinarians, veterinary technicians or other staff or volunteers who are certified by state agencies to treat wildlife. They follow professional codes and standards, and sometimes publish research in peer-reviewed journals.</p>
<p><div data-react-class="InstagramEmbed" data-react-props="{"url":"https://www.instagram.com/reel/CwNiHd5AkSL/?utm_source=ig_web_copy_link\u0026igshid=MzRlODBiNWFlZA==","accessToken":"127105130696839|b4b75090c9688d81dfd245afe6052f20"}"></div></p>
<h2>A growing data pool</h2>
<p>Until recently, most wildlife rehab records existed only in binders and file cabinets. As a result, studies drawing on these records typically used materials from a single location or focused on a particular species, such as bald eagles or foxes. </p>
<p>Recently, though, rehab centers have digitized hundreds of thousands of case records. Shareable digital records can improve wildlife conservation and public health. </p>
<p>For example, the <a href="https://www.wildlifecenter.org/">Wildlife Center of Virginia</a> has worked with government agencies and other rehab centers to establish the <a href="https://www.wild-one.org/">WILD-ONe database</a> as a tool for assessing trends in wildlife health. This will be an exciting area of research as more records are digitized and shared.</p>
<figure class="align-center zoomable">
<a href="https://images.theconversation.com/files/560206/original/file-20231117-19-6un51w.png?ixlib=rb-1.1.0&q=45&auto=format&w=1000&fit=clip"><img alt="Map showing distribution of wildlife centers that provided data for the study." src="https://images.theconversation.com/files/560206/original/file-20231117-19-6un51w.png?ixlib=rb-1.1.0&q=45&auto=format&w=754&fit=clip" srcset="https://images.theconversation.com/files/560206/original/file-20231117-19-6un51w.png?ixlib=rb-1.1.0&q=45&auto=format&w=600&h=348&fit=crop&dpr=1 600w, https://images.theconversation.com/files/560206/original/file-20231117-19-6un51w.png?ixlib=rb-1.1.0&q=30&auto=format&w=600&h=348&fit=crop&dpr=2 1200w, https://images.theconversation.com/files/560206/original/file-20231117-19-6un51w.png?ixlib=rb-1.1.0&q=15&auto=format&w=600&h=348&fit=crop&dpr=3 1800w, https://images.theconversation.com/files/560206/original/file-20231117-19-6un51w.png?ixlib=rb-1.1.0&q=45&auto=format&w=754&h=438&fit=crop&dpr=1 754w, https://images.theconversation.com/files/560206/original/file-20231117-19-6un51w.png?ixlib=rb-1.1.0&q=30&auto=format&w=754&h=438&fit=crop&dpr=2 1508w, https://images.theconversation.com/files/560206/original/file-20231117-19-6un51w.png?ixlib=rb-1.1.0&q=15&auto=format&w=754&h=438&fit=crop&dpr=3 2262w" sizes="(min-width: 1466px) 754px, (max-width: 599px) 100vw, (min-width: 600px) 600px, 237px"></a>
<figcaption>
<span class="caption">Locations in the U.S. and Canada where animals were found (blue dots) before being brought to wildlife rehabilitation centers (red stars) included in Miller et al., 2023.</span>
<span class="attribution"><a class="source" href="https://doi.org/10.1016/j.biocon.2023.110295">Miller et al., 2023</a>, <a class="license" href="http://creativecommons.org/licenses/by-nd/4.0/">CC BY-ND</a></span>
</figcaption>
</figure>
<h2>Threats vary by species</h2>
<p>Using this trove of data, we have been exploring patterns of wildlife health across North America. In our study, we <a href="https://doi.org/10.1016/j.biocon.2023.110295">identified key threats affecting wildlife</a> by region and for iconic and endangered species. </p>
<p>Overall, 12% of the animals brought to rehab centers during this period were harmed by vehicle collisions – the single largest cause of injury. For <a href="https://www.allaboutbirds.org/guide/Great_Horned_Owl/overview#">great horned owls</a>, which are common across the U.S., cars were the most common cause of admission – possibly because the owls commonly <a href="https://www.fws.gov/story/threats-birds-collisions-road-vehicles">forage at the same height as vehicles</a>, and may feed on road kill. </p>
<p>Other threats reflect various animals’ habitats and life patterns. Window collisions were the most common injury for the <a href="https://www.batcon.org/bat/eptesicus-fuscus/">big brown bat</a>, another species found in many habitats across the U.S. Fishing incidents were the main reason for admission of endangered <a href="https://www.fisheries.noaa.gov/species/kemps-ridley-turtle">Kemp’s ridley sea turtles</a>, which are found in the Gulf of Mexico and along the Atlantic coast.</p>
<p>Toxic substances and infectious diseases represented just 3.4% of cases, but were important for some species. <a href="https://www.allaboutbirds.org/guide/Bald_Eagle/overview">Bald eagles</a>, for example, were the species most commonly brought to centers with lead poisoning. Eagles and other raptors <a href="https://www.wildlifecenter.org/lead-toxicity-raptors">consume lead ammunition inadvertently</a> when they feed on carcasses left in the wild by hunters. </p>
<p>In southern Florida, hurricanes and floods resulted in spikes in the numbers of animals brought to rehab centers, reflecting the <a href="https://theconversation.com/scientists-at-work-sloshing-through-marshes-to-see-how-birds-survive-hurricanes-146067">impact of climate-driven extreme weather events on wildlife health</a>. </p>
<p>About one-third of animals in the cases we reviewed were successfully released back to the wild, though this varied greatly among species. For example, 68% of brown pelicans were released, but only 20% of bald eagles. Unfortunately, some 60% of the animals died from their injuries or illnesses, or had to be humanely euthanized because they were unable to recover.</p>
<h2>Spotlighting solutions</h2>
<p>Our results spotlight steps that can help conserve wildlife in the face of these threats. For example, transportation departments can build more <a href="https://interestingengineering.com/lists/29-of-the-most-heartwarming-wildlife-crossings-around-the-world">road crossings for wildlife</a>, such as bridges and underpasses, to help animals avoid being hit by cars.</p>
<figure class="align-center zoomable">
<a href="https://images.theconversation.com/files/560059/original/file-20231116-29-cl09dy.jpg?ixlib=rb-1.1.0&q=45&auto=format&w=1000&fit=clip"><img alt="A large wild cat emerges from an underpass beneath a highway." src="https://images.theconversation.com/files/560059/original/file-20231116-29-cl09dy.jpg?ixlib=rb-1.1.0&q=45&auto=format&w=754&fit=clip" srcset="https://images.theconversation.com/files/560059/original/file-20231116-29-cl09dy.jpg?ixlib=rb-1.1.0&q=45&auto=format&w=600&h=338&fit=crop&dpr=1 600w, https://images.theconversation.com/files/560059/original/file-20231116-29-cl09dy.jpg?ixlib=rb-1.1.0&q=30&auto=format&w=600&h=338&fit=crop&dpr=2 1200w, https://images.theconversation.com/files/560059/original/file-20231116-29-cl09dy.jpg?ixlib=rb-1.1.0&q=15&auto=format&w=600&h=338&fit=crop&dpr=3 1800w, https://images.theconversation.com/files/560059/original/file-20231116-29-cl09dy.jpg?ixlib=rb-1.1.0&q=45&auto=format&w=754&h=424&fit=crop&dpr=1 754w, https://images.theconversation.com/files/560059/original/file-20231116-29-cl09dy.jpg?ixlib=rb-1.1.0&q=30&auto=format&w=754&h=424&fit=crop&dpr=2 1508w, https://images.theconversation.com/files/560059/original/file-20231116-29-cl09dy.jpg?ixlib=rb-1.1.0&q=15&auto=format&w=754&h=424&fit=crop&dpr=3 2262w" sizes="(min-width: 1466px) 754px, (max-width: 599px) 100vw, (min-width: 600px) 600px, 237px"></a>
<figcaption>
<span class="caption">A mountain lion uses an underpass to safely traverse Route 97 near Bend, Oregon.</span>
<span class="attribution"><a class="license" href="http://creativecommons.org/licenses/by/4.0/">CC BY</a></span>
</figcaption>
</figure>
<p>Wildlife management agencies can <a href="https://doi.org/10.1007/s13280-018-1132-x">ban or limit use of ammunition and fishing gear that contain lead</a> to reduce lead poisoning. And governments can <a href="https://www.humanesociety.org/resources/wildlife-disaster-preparedness">incorporate wildlife into disaster management plans</a> to account for surges in wildlife rescues after extreme weather events.</p>
<p>People can also make changes on their own. They can drive more slowly and pay closer attention to wildlife crossing roads, switch their fishing and hunting gear to nonlead alternatives, and <a href="https://www.allaboutbirds.org/news/how-can-i-keep-birds-from-hitting-my-windows/">put decals or other visual indicators on windows</a> to reduce bat and bird collisions with the glass.</p>
<p>To learn more about animals in your area and ways to protect them, you can <a href="https://www.humanesociety.org/resources/how-find-wildlife-rehabilitator">visit or call your local wildlife rehab center</a>. You can also donate to these centers, which we believe do great work, and are often underfunded.</p>
<p>The scale of threats facing wild animals can seem overwhelming, but wildlife rehabbers show that helping one injured animal at a time can identify ways to save many more animal lives.</p><img src="https://counter.theconversation.com/content/214819/count.gif" alt="The Conversation" width="1" height="1" />
<p class="fine-print"><em><span>Tara K. Miller received funding from the National Science Foundation. </span></em></p><p class="fine-print"><em><span>Richard B. Primack does not work for, consult, own shares in or receive funding from any company or organization that would benefit from this article, and has disclosed no relevant affiliations beyond their academic appointment.</span></em></p>Hundreds of wildlife rehabilitation centers across the US and Canada treat sick and injured animals and birds. Digitizing their records is yielding valuable data on human-wildlife encounters.Tara K. Miller, Policy Research Specialist, Repair Lab, University of VirginiaRichard B. Primack, Professor of Biology, Boston UniversityLicensed as Creative Commons – attribution, no derivatives.tag:theconversation.com,2011:article/2176722023-11-17T03:26:59Z2023-11-17T03:26:59ZBig data play a huge role in US presidential elections. Do they have the same impact here?<p>A key reason Barack Obama won the 2012 US presidential election was his campaign’s use of “big data” to target specific voters. His team created multiple versions of ads aimed at niche audiences, taking care to test every message. </p>
<p>Naturally, some have worried about the potential power of these data-driven campaign techniques to manipulate voters. But have these methods taken over election campaigns in Australia? </p>
<p>In short, not really. Australian campaigns typically rely on much less data-intensive techniques due to a lack of resources, doubts about the data, and ethical and philosophical concerns about the approach.</p>
<p>I am a political scientist who studies political advertising in the United States, and I spent the first six months of 2023 in Australia as a Fulbright scholar. I interviewed campaign staff and political consultants about their use of various campaign techniques in state and federal elections. </p>
<p>My questions focused on political advertising – how it is targeted, the extent to which ads are tailored to specific audiences, and how campaigns test their messages. </p>
<p>So what do advertising campaigns look like? </p>
<p>First, while the expertise exists to do micro-targeting of individual voters based on sophisticated statistical modelling, most campaigns target broad categories of voters defined by their age, gender, where they live or the language they speak.</p>
<p>This kind of targeting, of course, has existed for decades. Campaigns have sent mail to specific addresses or knocked only on certain people’s doors.</p>
<p>Second, while a presidential campaign in the US might create tens of thousands of versions of an online political ad, such tailoring of ads to specific audiences is much more limited in Australia. </p>
<p>Third, ad testing relies heavily on the simple tools provided by Meta (owners of Facebook and Instagram) and focus groups. Large-scale testing done with online panellists is rare. </p>
<p>In short, most Australian campaigns do not resemble the data-intensive campaigns typical of presidential elections in the US. Why? </p>
<p>One reason is that campaigns do not have unlimited money and staff resources. At the end of the day, hiring a data scientist or creative staff to design ads for multiple audiences is a luxury most campaigns cannot afford. In contrast, more than US$6.6 billion (A$10.2 billion) was spent on the <a href="https://www.cnbc.com/2020/10/28/2020-election-spending-to-hit-nearly-14-billion-a-record.html">2020 presidential election</a>. </p>
<p>Second, campaign staff expressed some doubts about the data that do exist. While there was a lot of confidence in the voter roll provided by the Australian Electoral Commission, many interviewees reported that audience engagement on Facebook had declined considerably. In addition, it is now much more difficult to pinpoint where people spend their time because of privacy changes to Apple’s operating system. </p>
<p>Moreover, some campaigners, especially from the Greens, had ethical concerns about delivering different messages to different sets of voters. </p>
<p>Finally, there is a real disagreement about the wisdom of conducting a data-intensive campaign in which individual voters are targeted with tailored messages based on their beliefs, behaviours and demographic characteristics. Not only is this type of campaigning costly, but some argued that the key to winning an election is to send one broad message – or a small number of messages – to as many voters as possible. At the end of the day, parties want awareness of their candidates and an understanding of their central message. </p>
<hr>
<p>
<em>
<strong>
Read more:
<a href="https://theconversation.com/how-did-politicians-and-political-parties-get-my-mobile-number-and-how-is-that-legal-168750">How did politicians and political parties get my mobile number? And how is that legal?</a>
</strong>
</em>
</p>
<hr>
<p>So will Australian campaigns soon resemble the data-driven enterprises we see in the US? It seems unlikely. </p>
<p>First, in spite of public funding and few limits on raising money, Australian campaigns remain low-cost affairs compared with their US counterparts. </p>
<p>Second, the doubts about the effectiveness of data-driven microtargeted campaigns – and the data on which they rely – show no sign of abating. Indeed, one person who works for the Labor Party told me the party severely reduced the number of online ads it created between 2019 and 2022. The individual explained:</p>
<blockquote>
<p>In 2019, we created 1,000 different variations of digital ads, all informed by online experiments. We identified segments based on demography or geography, and we picked ads that did the best. But I’m not sure what value we got out of that hyper-optimisation – it was technological fetishisation. We didn’t stop to ask if it was a strategically intelligent campaign.</p>
</blockquote>
<p>Finally, while registered political parties in Australia are exempt from data privacy laws, that may not be the case forever if Australia follows Europe’s lead. New rules in the European Union restrict the use of sensitive personal data for micro-targeting political ads. </p>
<hr>
<p>
<em>
<strong>
Read more:
<a href="https://theconversation.com/australians-are-tired-of-lies-in-political-advertising-heres-how-it-can-be-fixed-189043">Australians are tired of lies in political advertising. Here's how it can be fixed</a>
</strong>
</em>
</p>
<hr>
<p>Earlier this year, the Australian Attorney-General’s Department released a <a href="https://www.ag.gov.au/integrity/consultations/review-privacy-act-1988#:%7E:text=The%20review%20covered%20areas%20including,for%20promoting%20good%20privacy%20practices">review of the Data Privacy Act of 1988</a>. Among the recommendations were limits on advertising targeting.</p>
<p>So worries about the potential of data-driven campaigns to manipulate Australian voters could prove to be more hype than reality.</p><img src="https://counter.theconversation.com/content/217672/count.gif" alt="The Conversation" width="1" height="1" />
<p class="fine-print"><em><span>Travis N. Ridout does not work for, consult, own shares in or receive funding from any company or organisation that would benefit from this article, and has disclosed no relevant affiliations beyond their academic appointment.</span></em></p>Targeted campaign advertising in Australia is very different from the United States. Here’s why.Travis N. Ridout, Professor of Government and Public Policy, Washington State UniversityLicensed as Creative Commons – attribution, no derivatives.tag:theconversation.com,2011:article/2155602023-10-13T12:17:30Z2023-10-13T12:17:30ZSouth Africa’s 2022 census missed 31% of people - big data could help in future<p>No census is ever exact: as academics Tom Moultrie and Rob Dorrington at the University of Cape Town have <a href="https://www.un.org/development/desa/pd/sites/www.un.org.development.desa.pd/files/files/documents/2020/Oct/undesa_pd_tp_2020_tp_population_estimates.pdf#page=8">noted previously</a>:</p>
<blockquote>
<p>a census is not, in reality, a full and accurate count of the number of people in a country; rather, it is itself an estimate of the size of the population at a moment in time. </p>
</blockquote>
<p>South Africa has announced the results of its fourth census as a democracy – <a href="https://www.statssa.gov.za/?p=16711">Census 2022</a>. I have been involved in the process for the last four years as chair of South Africa’s National Statistics Council. As outgoing chair, my last task was to <a href="https://www.statssa.gov.za/?p=15192">take part</a> in the release of Census 2022.</p>
<p>The census found that the national population has grown to <a href="https://www.statssa.gov.za/?p=16711">62 million</a>, up 10.3 million from the last census <a href="https://www.statssa.gov.za/publications/P03014/P030142011.pdf">in 2011</a>. Gauteng is now clearly the most populous province in the country, with 15.1 million people, overtaking KwaZulu-Natal (12.4 million). The Western Cape jumped from fifth to being the third largest province, with 7.4 million people. These figures are important because they inform resource allocation by government.</p>
<p>What is perhaps most striking about Census 2022 is the very high <a href="https://www.statssa.gov.za/publications/P03015/P030152022.pdf#page=11">undercount</a> – 31% of people and 30% of households were missed (or chose not to self-enumerate, either online or via zero-rated telephone methods). This is the highest undercount of any post-apartheid census; sadly, it may set a new international record. </p>
<p>A census is immediately followed by a <a href="https://www.statssa.gov.za/publications/P03015/P030152022.pdf">Post Enumeration Survey</a>, which identifies where the census missed people. This allows Statistics SA to develop adjustment factors, or weights, so that the final data represents an adjusted final tally. The Post Enumeration Survey is used to manage the undercount. Census undercounts are the norm, <a href="https://www.un.org/development/desa/pd/sites/www.un.org.development.desa.pd/files/files/documents/2020/Oct/undesa_pd_tp_2020_tp_population_estimates.pdf#page=8">not the exception</a>. But it is safe to assume that with weighting on this scale – adjusting for an undercount of 31.06% – analysts may identify some confounding results. </p>
<p>At aggregate level, <a href="https://census.statssa.gov.za/#/">Census 2022</a> is robust. At sub-national – and especially sub-provincial – levels, however, it may be less so. Only time and data analysis will tell.</p>
<p>The census confirmed the global trend of <a href="https://www.icf.com/insights/health/declining-survey-response-rate-problem">declining survey response rates</a>. People are less and less inclined to be involved in the process. This raises the question: does a fieldwork-based census have a future? Given the challenges that faced Census 2022, I believe the census may need to be re-imagined as a very different exercise. This requires <a href="https://www.statssa.gov.za/">Statistics South Africa</a>, which conducts the census, to fully engage with big data to bring the process into the 21st century.</p>
<h2>The process</h2>
<p>South Africa’s National Statistics Council, an independent body of experts that advises the statistician-general and the minister in the presidency regarding statistics, had secured a number of local and international experts – as had Stats SA – to stress test the census and the Post Enumeration Survey. Council never has prior sight of the data: its job is to focus on methods and process.</p>
<p>The experts do engage with the data and flagged only a few variables (mortality data, and some service and asset questions which had too many non-responses to be reliable) as requiring a cautionary note. Council engaged vigorously with the experts and Stats SA, and with no red flag raised by any, we declared the census <a href="https://census.statssa.gov.za/#/">“fit for purpose”</a>.</p>
<p>It is notable that Stats SA routinely conducts a post enumeration survey. Many countries do not, even when there is systematic undercounting of particular groups (often young men, children and minorities). Moreover, Stats SA will make available both the weighted and the raw data for analysts to examine in detail. This transparency should be welcomed, given that (as previously noted by the <a href="https://unstats.un.org/unsd/demographic/sources/census/wphc/QA.htm">United Nations Statistics Division</a>) issues of undercounting affect all countries, and estimating the undercount and whether to adjust the data is a political issue “throughout the world”. The undercount was high, but not as a result of any lack of effort or commitment from Stats SA.</p>
<h2>Why the undercount</h2>
<p>The undercount is the result of many factors. </p>
<p>First, the context matters. This time round it was as bad as it could be, with the COVID-19 pandemic affecting training and supply chains for equipment. The pandemic also generated anxiety in a populace that had been avoiding contact with strangers as part of social distancing. Census planning usually starts three or four years prior to fieldwork. Training about 100,000 enumerators is a major effort in its own right, combined with the shift to digital platforms for the first time. All were affected by the pandemic.</p>
<p>The fieldwork took place after the devastating <a href="https://theconversation.com/south-africas-deadly-july-2021-riots-may-recur-if-theres-no-change-186397">July 2021 insurrection</a>, and after the hard-fought <a href="https://www.elections.org.za/pw/Elections-and-results/Municipal-Elections-2021">local elections</a>. The process also coincided with xenophobic violence meted out by the anti-migrant pressure group-<a href="https://www.theguardian.com/global-development/2023/sep/26/south-africa-anti-migrant-vigilante-operation-dudula-registers-as-party-2024-elections">turned-political party</a> <a href="https://theconversation.com/rising-vigilantism-south-africa-is-reaping-the-fruits-of-misrule-179891">Operation Dudula</a> in Johannesburg. Taken together, the effect was a deep-seated reluctance to open doors to strangers, particularly those asking lots of questions.</p>
<p>A second factor that affected the gathering of data was the fact that there is <a href="https://www.afrobarometer.org/wp-content/uploads/migrated/files/publications/Dispatches/ad474-south_africans_trust_in_institutions_reaches_new_low-afrobarometer-20aug21.pdf">very low trust in the government</a>. Although the census is conducted by Stats SA, which is an independent entity, it is seen as “government”. This label didn’t make it easy to persuade people to allow an enumerator into their dwellings and answer questions. </p>
<p>People in the Western Cape, the only province not run by the African National Congress, were particularly resistant to being enumerated or self-enumerating. This was true even after the provincial premier and Cape Town mayor made public calls for people to comply. The undercount in the Western Cape stands at 35.58% of people and 36.3% of households. In the Free State, by comparison, the undercount is 20.95% of people and 17.93% of households.</p>
<p>A third factor was that response rates have been getting consistently lower over at least the last decade. This has been true for Stats SA and other entities undertaking primary research. The decision to go digital was an attempt to open different avenues for people to complete the questionnaire online, or by phone, to improve response rates.</p>
<p>People appear to be sick and tired of being polled by everyone, from their local supermarket to endless tele-marketers and others. They also appear much more wary of sharing their data. What, then, is the future for the census?</p>
<h2>Enter big data</h2>
<p>Countries around the world are facing the same challenge of low response rates. </p>
<p>The advent of big data opens intriguing possibilities. </p>
<p>A first step would be to harvest data from the records kept by government departments (assuming they are run well). In addition, data could be unlocked if a working relationship was developed with private sector entities, such as suppliers and banks. </p>
<p>Becoming far more tech-savvy, and encouraging people to engage with Stats SA digitally, could be combined with other options to compile a national population dataset. It would also represent a significant cost-saving. This approach – harvesting data rather than gathering it directly – is being considered by many countries, but has not yet been attempted, and Stats SA needs to carefully consider this option.</p>
<p>Stats SA needs to fully engage with the world of big data, and the key players in that data ecosystem. It has convening authority, and should be engaging all key players, whether they are academic, private sector or others. </p>
<p>At the very least, an alternative way of conducting the next census in 2032 must be rigorously examined and tested. </p>
<p>Big data is not the answer to all the challenges that faced Census 2022, but it may be a key enabler for gathering reliable national data in the future.</p><img src="https://counter.theconversation.com/content/215560/count.gif" alt="The Conversation" width="1" height="1" />
<p class="fine-print"><em><span>David Everatt does not work for, consult, own shares in or receive funding from any company or organisation that would benefit from this article, and has disclosed no relevant affiliations beyond their academic appointment.</span></em></p>Big data is not the answer to all the challenges that faced Census 2022, but it may be a key enabler for gathering reliable national data in the future.David Everatt, Professor of Urban Governance, University of the WitwatersrandLicensed as Creative Commons – attribution, no derivatives.tag:theconversation.com,2011:article/2110842023-08-25T13:39:37Z2023-08-25T13:39:37ZSetting the stage for a better understanding of complex brain disorders<p>We often compare the brain to a machine with wheels, cogs, and belts. In this analogy, when something breaks, the entire mechanism skips a beat or grinds to a halt. However, more often than not this isn’t what happens with our brains. Instead, they’re more like a theatre. Here, neurons are the musicians, actors, and dancers, and they improvise a performance that shapes our thoughts and lives.</p>
<p>I’m an electronic and computer engineer at the <a href="https://www.epu.ntua.gr/">DSS Lab</a> of the National Technical University of Athens. In December 2019, Ioannis Stavropoulos, a neuroscientist at King’s College London, introduced me to his colleague Elissaios Karageorgiou of the Neurological Institute of Athens. They wanted to talk about an idea they had about neurology and, in a way, theatre, over coffee.</p>
<p>In any piece of theatre, mistakes happen – a violin might miss a note, a drummer could skip a beat, an actor might muddle a line or a dancer stumble. Sometimes, many things go wrong simultaneously, and the audience is left wondering what’s going on. Was it the singer who was off? Was it the pianist who hit the wrong chord? Did the lights go off at the wrong time and confuse them both?</p>
<p>Complex brain disorders (CoBraD) are very much like that. These include Alzheimer’s disease, sleep problems, and epilepsy. We see their symptoms as performance missteps, yet it’s hard to group, label, and know their causes. When multiple symptoms show up at once, diagnosis becomes particularly challenging.</p>
<p>It would be hard to know what’s wrong in a musical or play by just listening to a second or two every half hour. Similarly, it’s tougher to diagnose a medical problem if we only check the patient briefly, such as during occasional doctor visits. As a result, CoBraDs may stay under- or undiagnosed for a long time or be misdiagnosed. More than one can exist at the same time, and the diagnosis and treatment are expensive, at times inaccessible for patients, and often ineffective.</p>
<h2>Paying attention to the whole performance</h2>
<p>Much like how a piece of theatre relies on each artist to play her or his part well for a captivating performance, diagnosing CoBraDs demands a wide range of accurate and harmonised data. Alzheimer’s, sleep disorders, and epilepsy are among the conditions that Ioannis and Elissaios are studying and treating. Recognising the limitations of traditional diagnostic methods, they turned their attention to real-world data (RWD), meaning data collected directly from patients not taking part in a clinical trial.</p>
<p>Data gathered from <a href="https://www.who.int/health-topics/clinical-trials">clinical trials</a> are more reliable than real-world data – they’re the result of experiments performed in strict and controlled conditions. However, they are often hard and expensive to get, limited in size, and they may not fully represent the complexity and variability of the real world.</p>
<p>In contrast, real-world data encompass a wider array of information sources, from electronic health records and patient visits to medical devices like MRIs and wearables. When aggregated, these diverse data points become “big data”, offering a more comprehensive view of patient health. This holistic approach can reveal patterns and insights that might be missed in more conventional, narrower diagnostic methods.</p>
<p>Collecting significant amounts of real-world data is just the beginning. The real challenge lies in harmonising and analysing it all to extract meaningful insights and then finding ways to use them to diagnose and treat patients. To achieve this, we sought expertise from various scientific disciplines. What became our vision was to create a digital platform where neuroscientists could store and share large amounts of data, analyse them, and use them to devise new diagnostic processes and criteria that would be more complex and nuanced than what human clinicians can handle.</p>
<p>These processes would be built into the platform to support clinicians in making decisions for their patients when diagnosing or treating them. These are called decision support systems, and when they use tools like artificial intelligence, enhancing the competences of human experts in a technical or scientific field, they’re called expert systems.</p>
<p>Scientists have proposed numerous ideas hinting that faint clues and varied signals might point to early detection of CoBraD. Many remain unproven, and some are tough to track without computers. For instance, slight changes in sleep, coupled with specific MRI signs, could suggest an early brain disorder. Rather than waiting years for clear symptoms to emerge, doctors could act swiftly, improving the patient’s prospects.</p>
<p>This is how the idea for the <a href="https://www.mes-cobrad.eu/">Multidisciplinary Expert System for the Assessment and Management of Complex Brain Disorders</a> (MES-CoBraD) was born. Bringing together with experts in medicine, engineering, and computer science, we are building a software platform and performing medical research using it.</p>
<h2>When doctors, practitioners, software engineers and AI work together</h2>
<p>MES-CoBraD evolved into an EU-funded project that now includes 14 universities, companies, and hospitals across Europe. The underlying concept is straightforward: data and observations produced by clinical practice are used by medical research to enhance that very practice.</p>
<p>This continuous circular collaboration can use technology as a link and an enabler. Researchers and clinicians collect and anonymise patient data and upload them to the platform. The researchers form scientific hypotheses, analyse the data, train AI models, and test their hypotheses.</p>
<p>Should they achieve a breakthrough, clinicians would be able to directly use the platform’s algorithms to diagnose patients and provide treatments. Related data would in turn will be anonymised and serve to test new hypotheses, aid new statistical analyses, train AI models, or refine existing ones.</p>
<p>The challenges are many. When designing new experiments, we must ensure our data are unbiased. We are also investigating and addressing the ethical implications of using artificial intelligence in medicine. For instance, how do we guarantee that its suggestions are understood by the clinician and can be explained to the patient? How can we be certain that they don’t inadvertently favour one patient group over another, or prioritise cost savings over human life? If AI makes a mistake, who takes responsibility?</p>
<p>In MES-CoBraD, we have been venturing into some uncharted territory, but always with specific goals in mind. Although the platform is being made to work in multiple medical fields, the focus now is finding ways of using a very detailed picture of patients’ health (a process called deep phenotyping) in conjunction with advanced analytics tools and AI to diagnose and manage CoBraDs. In essence, we aim to simultaneously tune the instruments, hone the actors’ lines, and adjust the score.</p>
<h2>A personal turn on the stage</h2>
<p>Interestingly, my own life’s “play” has its moments of dissonance, as I sometimes wake in the middle of the night and find it hard to get back to sleep. I don’t think it’s anything serious, but as a scientist and longtime researcher, I would never miss the chance to test our own methods and processes on myself. I thus signed up as a test subject for our sleep study and wore a device called <a href="https://www.sciencedirect.com/topics/medicine-and-dentistry/actigraphy">“actigraph”</a> on my wrist for a week, and it kept track of my activities, diet, and sleep. I took memory tests and answered questions, and finally there was the main event: I got hooked up with around 40 or 50 cables, tubes and sensors, and slept in the clinic for the night. As a bonus, I offered months of smartwatch health data, which was anonymised and included in the platform.</p>
<p>I am happy to report that for now my sleep issues are likely to be stress-related. However, if deep phenotyping and AI end up diagnosing something worse – say, an early onset complex brain disorder – the question for someone in my position would be: “Should I worry or celebrate the scientific breakthrough if it does?”</p>
<hr>
<p><em>This article is the result of The Conversation’s collaboration with <a href="https://ec.europa.eu/research-and-innovation/en/horizon-magazine">Horizon</a>, the EU research and innovation magazine. In June the magazine published an <a href="https://ec.europa.eu/research-and-innovation/en/horizon-magazine/brain-disorders-trigger-search-new-clues-and-cures">interview with the authors about their research</a>.</em></p><img src="https://counter.theconversation.com/content/211084/count.gif" alt="The Conversation" width="1" height="1" />
<p class="fine-print"><em><span>Christos Ntanos is the Project Coordinator of the "MES-CoBraD -- Multidisciplinary Expert System for the Assessment & Management of Complex Brain Disorders" project, which has received funding from the European Union's Horizon 2020 research and innovation programme under grant agreement No 965422.</span></em></p>Disorders such as Alzheimer’s and epilepsy are difficult to diagnose with only occasional doctor visits. A new approach would allow fathering of extensive real-world data directly from patients.Christos Ntanos, Research director at the Decision Support Systems (DSS) Laboratory, National Technical University of AthensLicensed as Creative Commons – attribution, no derivatives.tag:theconversation.com,2011:article/2115772023-08-15T19:31:37Z2023-08-15T19:31:37ZZoom’s scrapped proposal to mine user data causes concern about our virtual and private Indigenous Knowledge<figure><img src="https://images.theconversation.com/files/542670/original/file-20230814-17-i0xdd7.jpg?ixlib=rb-1.1.0&rect=131%2C108%2C5044%2C3771&q=45&auto=format&w=496&fit=clip" /><figcaption><span class="caption">Platforms like Zoom have been helpful in bridging geographical distances. However, a recent proposal to mine data raises questions about ownership of Indigenous Knowledge. </span> <span class="attribution"><span class="source">(Chris Montgomery/Unsplash)</span></span></figcaption></figure><iframe style="width: 100%; height: 100px; border: none; position: relative; z-index: 1;" allowtransparency="" allow="clipboard-read; clipboard-write" src="https://narrations.ad-auris.com/widget/the-conversation-canada/zooms-scrapped-proposal-to-mine-user-data-causes-concern-about-our-virtual-and-private-indigenous-knowledge" width="100%" height="400"></iframe>
<p>As reported on Aug. 6, <a href="https://stackdiary.com/zoom-terms-now-allow-training-ai-on-user-content-with-no-opt-out/">Zoom recently attempted to rewrite its Terms of Service with ambiguous language that would permit the extraction of user data for the purpose of training AI</a>. </p>
<p>However, after <a href="https://variety.com/2023/digital/news/zoom-response-customer-content-ai-training-1235689725/">public pushback</a>, <a href="https://blog.zoom.us/zooms-term-service-ai/">Zoom began to rectify that clause the very next day</a>, fully committing to a “no AI training” set of policies by Aug. 11. </p>
<p>Even though Zoom pedalled back this time, their drive to gather data highlights the possibility of future hidden data extraction by them and other big tech companies.</p>
<p>More specifically, as a researcher working with and looking at Indigenous communities and their data, I am concerned about the privacy of these valuable data sets from Indigenous communities on Turtle Island. </p>
<h2>Vulnerable Indigenous Knowledge</h2>
<p>Over the past three years, Zoom calls have become a tool for organization and activism for many Indigenous communities. </p>
<p>For my own work, I use video and voice chat which lets us balance geographical differences to collaborate and share, as well as access communities that are hard to reach. Discussing issues with queer community members of different Indigenous Nations is often private and perhaps even sacred. </p>
<p>These conversations have elements that are public facing, but they also contain wisdom from Elders or Knowledge Keepers specifically trained to know what they can and cannot share in specific spaces. Some of this knowledge is sacred and is part of promoting and preserving Indigenous (and sometimes queer) ways of being. </p>
<h2>A valuable commodity</h2>
<p>This private information is constantly at risk of extraction from companies seeking to monetize or otherwise gain from our data.</p>
<p>Indigenous Knowledge represents a large gap in current big data. AI only works with large data sets which enables predictive technology to operate. </p>
<p>With knowledges that are primarily oral, it is difficult to gather proper data sets that often come from writing. The possibility for big companies to gather audio and visual data, could render this oral information visible by machines. </p>
<h2>Protecting communities</h2>
<p><a href="https://static1.squarespace.com/static/557744ffe4b013bae3b7af63/t/557f2ee5e4b0220eff4ae4b5/1434398437409/Tuck+and+Yang+R+Words_Refusing+Research.pdf">“Refusing research”</a> has been an important concept for protecting marginalized communities from the extractive practices of researchers aiming to obtain data.</p>
<p>However, if platforms are extracting data without our knowledge, or demand our consent in order to use a service, a conflict emerges.</p>
<p>The conflict becomes one of <a href="https://techcrunch.com/2023/08/08/zoom-data-mining-for-ai-terms-gdpr-eprivacy/">free choice versus free-to-leave</a>: If we do not consent to use the infrastructure, we simply do not get access to that service. Access to voice and video sharing infrastructure has been a fundamental component of activism and community research, especially post COVID-19.</p>
<h2>Can we ‘opt-out?’</h2>
<p>Can we accept or refuse to be turned into research data?</p>
<p>Even though there is a permissions element, organizations are often gathering our data in exchange for using their services. For example, Fitbit <a href="https://www.fitbit.com/global/en-ca/legal/privacy-policy">gathers massive amounts of health data from users (with permission) that can be used to train AI</a>. </p>
<p>Each individual who is opting for nearly any big service is being tracked to some capacity. And so, there needs to be a critical element of what is considered private. </p>
<p>Likewise, Zoom has the ability to gather this data, whether or not they use it for AI with consent. There is an anxiety that next time, the ambiguity will go unnoticed or perhaps force consent to access a seemingly necessary service.</p>
<p>As someone who looks at ethical data collection and mobilization, I believe we all need to be critical of those requests to have access to our private data when using these services.</p>
<figure class="align-center ">
<img alt="A woman looks at a computer screen monitor" src="https://images.theconversation.com/files/542709/original/file-20230815-27813-p3choq.jpg?ixlib=rb-1.1.0&q=45&auto=format&w=754&fit=clip" srcset="https://images.theconversation.com/files/542709/original/file-20230815-27813-p3choq.jpg?ixlib=rb-1.1.0&q=45&auto=format&w=600&h=400&fit=crop&dpr=1 600w, https://images.theconversation.com/files/542709/original/file-20230815-27813-p3choq.jpg?ixlib=rb-1.1.0&q=30&auto=format&w=600&h=400&fit=crop&dpr=2 1200w, https://images.theconversation.com/files/542709/original/file-20230815-27813-p3choq.jpg?ixlib=rb-1.1.0&q=15&auto=format&w=600&h=400&fit=crop&dpr=3 1800w, https://images.theconversation.com/files/542709/original/file-20230815-27813-p3choq.jpg?ixlib=rb-1.1.0&q=45&auto=format&w=754&h=503&fit=crop&dpr=1 754w, https://images.theconversation.com/files/542709/original/file-20230815-27813-p3choq.jpg?ixlib=rb-1.1.0&q=30&auto=format&w=754&h=503&fit=crop&dpr=2 1508w, https://images.theconversation.com/files/542709/original/file-20230815-27813-p3choq.jpg?ixlib=rb-1.1.0&q=15&auto=format&w=754&h=503&fit=crop&dpr=3 2262w" sizes="(min-width: 1466px) 754px, (max-width: 599px) 100vw, (min-width: 600px) 600px, 237px">
<figcaption>
<span class="caption">In the future, will we find ourselves agreeing to give up our data just to use video platform software like Zoom?</span>
<span class="attribution"><span class="source">(Unsplash)</span></span>
</figcaption>
</figure>
<h2>Crucial access to data</h2>
<p>The relationship between data and Indigenous communities and the Canadian government has always been fraught. However, after the work of the <a href="https://www.rcaanc-cirnac.gc.ca/eng/1450124405592/1529106060525">Truth and Reconciliation Commission</a> in Canada (which concluded in 2015), it became even more clear that access to data and information is crucial to achieving justice and truth in relation to our histories.</p>
<p>For Indigenous peoples whose history has been systematically erased, demanding that organizations return records and data has become an important element of achieving the truth behind the experiences of Indian Residential School survivors. Communities have both the desire and need to have their data returned so that they can maintain <a href="https://fnigc.ca/ocap-training">ownership, control, access and manage permissions</a> to access information.</p>
<h2>Ease of Zoom for communication</h2>
<p>In-person collaboration between Indigenous communities can be difficult because of things like geographical differences, the lack of public transportation, and interruptions in Indigenous sovereignty. These issues continue the social and political fragmentation caused by settler colonialism to isolate these communities from one another.</p>
<p>Many of these challenges have been alleviated by information technologies like Zoom. And a platform like Zoom has been potentially unifying by bridging space. However, it could also become a tool to recreate the problem of data extraction in a new way. </p>
<p>We need to be attentive to these kinds of data gathering possibilities that offer to extract data from users. </p>
<p>These technological infrastructures may disproportionately harm Indigenous communities by making their private and sacred knowledges legible by AI. Data collection for AI could lead to the commodification of this sacred knowledge for profit. </p>
<p>Protecting this kind of data is not just the responsibility of Indigenous communities but a shared commitment that has a present and future urgency.</p><img src="https://counter.theconversation.com/content/211577/count.gif" alt="The Conversation" width="1" height="1" />
<p class="fine-print"><em><span>Andrew Wiebe does not work for, consult, own shares in or receive funding from any company or organisation that would benefit from this article, and has disclosed no relevant affiliations beyond their academic appointment.</span></em></p>In-person collaboration between Indigenous communities has been aided by information technologies like Zoom. However, recent attempts to mine personal data raise concerns about data ownership.Andrew Wiebe, PhD Student, Information, University of TorontoLicensed as Creative Commons – attribution, no derivatives.tag:theconversation.com,2011:article/2065742023-06-04T20:04:48Z2023-06-04T20:04:48ZSocial media snaps map the sweep of Japan’s cherry blossom season in unprecedented detail<figure><img src="https://images.theconversation.com/files/528956/original/file-20230530-19-j7i2pi.jpg?ixlib=rb-1.1.0&rect=0%2C0%2C3456%2C3559&q=45&auto=format&w=496&fit=clip" /><figcaption><span class="caption">
</span> <span class="attribution"><a class="source" href="https://unsplash.com/photos/-6yI4Z8ehlU">Kazuo Ota / Unsplash</a></span></figcaption></figure><p>Social media contains enormous amounts of data about people, our everyday lives, and our interactions with our surroundings. As a byproduct, it also contains a vast trove of information about the natural world.</p>
<p>In a <a href="https://www.sciencedirect.com/science/article/abs/pii/S0367253023001019#sec0024a">new study published in Flora</a>, we show how social media can be used for “incidental citizen science”. From photos posted to a social site, we mapped countrywide patterns in nature over a decade in relatively fine detail.</p>
<p>Our case study was the annual spread of cherry blossom flowering across Japan, where millions of people view the blooming each year in a cultural event called “hanami”. The flowering spreads across Japan in a wave (“<a href="https://en.wikipedia.org/wiki/Cherry_blossom_front">sakura zensen</a>” or 桜前線) following the warmth of the arriving spring season.</p>
<figure class="align-center ">
<img alt="ALT TEXT" src="https://images.theconversation.com/files/529000/original/file-20230530-15-mix84k.jpeg?ixlib=rb-1.1.0&q=45&auto=format&w=754&fit=clip" srcset="https://images.theconversation.com/files/529000/original/file-20230530-15-mix84k.jpeg?ixlib=rb-1.1.0&q=45&auto=format&w=600&h=399&fit=crop&dpr=1 600w, https://images.theconversation.com/files/529000/original/file-20230530-15-mix84k.jpeg?ixlib=rb-1.1.0&q=30&auto=format&w=600&h=399&fit=crop&dpr=2 1200w, https://images.theconversation.com/files/529000/original/file-20230530-15-mix84k.jpeg?ixlib=rb-1.1.0&q=15&auto=format&w=600&h=399&fit=crop&dpr=3 1800w, https://images.theconversation.com/files/529000/original/file-20230530-15-mix84k.jpeg?ixlib=rb-1.1.0&q=45&auto=format&w=754&h=501&fit=crop&dpr=1 754w, https://images.theconversation.com/files/529000/original/file-20230530-15-mix84k.jpeg?ixlib=rb-1.1.0&q=30&auto=format&w=754&h=501&fit=crop&dpr=2 1508w, https://images.theconversation.com/files/529000/original/file-20230530-15-mix84k.jpeg?ixlib=rb-1.1.0&q=15&auto=format&w=754&h=501&fit=crop&dpr=3 2262w" sizes="(min-width: 1466px) 754px, (max-width: 599px) 100vw, (min-width: 600px) 600px, 237px">
<figcaption>
<span class="caption">Celebrating the cherry blossom is a centuries-old tradition in Japan.</span>
<span class="attribution"><span class="source">Shutterstock</span></span>
</figcaption>
</figure>
<p>The <a href="https://en.wikipedia.org/wiki/Hanami">hanami festival</a> has been documented for centuries, and research shows climate change is making <a href="https://iopscience.iop.org/article/10.1088/1748-9326/ac6bb4">early blossoming more likely</a>. The advent of mobile phones – and social network sites that allow people to <a href="https://www.sciencedirect.com/science/article/abs/pii/S1574954116302321">upload photos tagged with time and location data</a> – presents a new opportunity to study how Japan’s flowering events are affected by seasonal climate. </p>
<h2>Why are flowers useful to understand how nature is being altered by climate change?</h2>
<p>Many flowering plants, including the cherry blossoms of Japan (<em>Prunus</em> subgenus <em>Cerasus</em>), require insect pollination. To reproduce, plant flowers bloom at optimal times to receive visits from insects like bees. </p>
<p>Temperature is <a href="https://journals.plos.org/plosone/article?id=10.1371/journal.pone.0200549">an important mechanism</a> for plants to trigger this flowering. <a href="https://onlinelibrary.wiley.com/doi/epdf/10.1111/j.1461-0248.2008.01269.x">Previous research</a> has highlighted how climate change may create mismatches in space or time between the blooming of plants and the emergence of pollinating insects.</p>
<p><iframe id="rtiQ0" class="tc-infographic-datawrapper" src="https://datawrapper.dwcdn.net/rtiQ0/2/" height="400px" width="100%" style="border: none" frameborder="0"></iframe></p>
<p>It has been difficult for researchers to map the extent of this problem in detail, as its study requires simultaneous data collection over large areas. The use of citizen science images deliberately, or incidentally, uploaded to social network sites enables <a href="https://en.wikipedia.org/wiki/Big_data">big data</a> solutions.</p>
<h2>How did we conduct our study?</h2>
<p>We collected images from Japan uploaded to <a href="https://www.flickr.com/">Flickr</a> between 2008 and 2018 that were tagged by users as “cherry blossoms”. We used computer vision techniques to analyse these images, and to provide sets of keywords describing their image content. </p>
<p>Next, we automatically filtered out images appearing to contain content that the computer vision algorithms determined didn’t match our targeted cherry blossoms. For instance, many contained images of autumn leaves, another popular ecological event to view in Japan. </p>
<p>The locations and timestamps of the remaining cherry blossom images were then used to generate marks on a map of Japan showing the seasonal wave of sakura blossoms, and to estimate peak bloom times each year in different cities.</p>
<h2>Checking the data</h2>
<p>An important component of any scientific investigation is validation – how well does a proposed solution or data set represent the real-world phenomenon under study? </p>
<figure class="align-right zoomable">
<a href="https://images.theconversation.com/files/528702/original/file-20230528-21-4fxpkv.png?ixlib=rb-1.1.0&q=45&auto=format&w=1000&fit=clip"><img alt="" src="https://images.theconversation.com/files/528702/original/file-20230528-21-4fxpkv.png?ixlib=rb-1.1.0&q=45&auto=format&w=237&fit=clip" srcset="https://images.theconversation.com/files/528702/original/file-20230528-21-4fxpkv.png?ixlib=rb-1.1.0&q=45&auto=format&w=600&h=591&fit=crop&dpr=1 600w, https://images.theconversation.com/files/528702/original/file-20230528-21-4fxpkv.png?ixlib=rb-1.1.0&q=30&auto=format&w=600&h=591&fit=crop&dpr=2 1200w, https://images.theconversation.com/files/528702/original/file-20230528-21-4fxpkv.png?ixlib=rb-1.1.0&q=15&auto=format&w=600&h=591&fit=crop&dpr=3 1800w, https://images.theconversation.com/files/528702/original/file-20230528-21-4fxpkv.png?ixlib=rb-1.1.0&q=45&auto=format&w=754&h=743&fit=crop&dpr=1 754w, https://images.theconversation.com/files/528702/original/file-20230528-21-4fxpkv.png?ixlib=rb-1.1.0&q=30&auto=format&w=754&h=743&fit=crop&dpr=2 1508w, https://images.theconversation.com/files/528702/original/file-20230528-21-4fxpkv.png?ixlib=rb-1.1.0&q=15&auto=format&w=754&h=743&fit=crop&dpr=3 2262w" sizes="(min-width: 1466px) 754px, (max-width: 599px) 100vw, (min-width: 600px) 600px, 237px"></a>
<figcaption>
<span class="caption">Blossom dates calculated from social media images compare well with official data.</span>
<span class="attribution"><span class="source">ElQadi et al.</span>, <span class="license">Author provided</span></span>
</figcaption>
</figure>
<p>Our study using social network site images was validated against the detailed information published by the <a href="https://www.japan.travel/en/see-and-do/cherry-blossom-forecast-2023/">Japan National Tourism Organization</a>. </p>
<p>We also manually examined a subset of images to confirm the presence of cherry flowers. </p>
<p>Plum flowers (<em>Prunus mume</em>) look very similar to cherry blossoms, especially to tourists, and they are frequently mistaken and mislabelled as cherry blossoms. We used visible “notches” at the end of cherry petals, and other characteristics, to distinguish cherries from plums.</p>
<p>Taken together, the data let us map the flowering event as it unfolds across Japan. </p>
<figure class="align-center zoomable">
<a href="https://images.theconversation.com/files/528769/original/file-20230529-17-wmgf5g.gif?ixlib=rb-1.1.0&q=45&auto=format&w=1000&fit=clip"><img alt="An animated map showing cherry blossom flowering across Japan" src="https://images.theconversation.com/files/528769/original/file-20230529-17-wmgf5g.gif?ixlib=rb-1.1.0&q=45&auto=format&w=754&fit=clip" srcset="https://images.theconversation.com/files/528769/original/file-20230529-17-wmgf5g.gif?ixlib=rb-1.1.0&q=45&auto=format&w=600&h=619&fit=crop&dpr=1 600w, https://images.theconversation.com/files/528769/original/file-20230529-17-wmgf5g.gif?ixlib=rb-1.1.0&q=30&auto=format&w=600&h=619&fit=crop&dpr=2 1200w, https://images.theconversation.com/files/528769/original/file-20230529-17-wmgf5g.gif?ixlib=rb-1.1.0&q=15&auto=format&w=600&h=619&fit=crop&dpr=3 1800w, https://images.theconversation.com/files/528769/original/file-20230529-17-wmgf5g.gif?ixlib=rb-1.1.0&q=45&auto=format&w=754&h=777&fit=crop&dpr=1 754w, https://images.theconversation.com/files/528769/original/file-20230529-17-wmgf5g.gif?ixlib=rb-1.1.0&q=30&auto=format&w=754&h=777&fit=crop&dpr=2 1508w, https://images.theconversation.com/files/528769/original/file-20230529-17-wmgf5g.gif?ixlib=rb-1.1.0&q=15&auto=format&w=754&h=777&fit=crop&dpr=3 2262w" sizes="(min-width: 1466px) 754px, (max-width: 599px) 100vw, (min-width: 600px) 600px, 237px"></a>
<figcaption>
<span class="caption">Images uploaded to social media over a ten year period 2008-2018, let us map the cherry blossom front as it sweeps across Japan.</span>
<span class="attribution"><span class="source">ElQadi et al.</span>, <span class="license">Author provided</span></span>
</figcaption>
</figure>
<h2>Out-of-season blooms</h2>
<p>Our social network site analysis was sufficiently detailed to accurately pinpoint the annual peak spring bloom in the major cities of <a href="https://en.wikipedia.org/wiki/Tokyo">Tokyo</a> and <a href="https://en.wikipedia.org/wiki/Kyoto">Kyoto</a>, to within a few days of official records. </p>
<p>Our data also revealed the presence of a consistent, and persistent, out-of-season cherry bloom in autumn. Upon further searching, we discovered that this “unexpected” seasonal bloom had also been noted in <a href="https://www.bbc.com/news/world-asia-45898333">mainstream media</a> in recent years. We thus confirmed that this is a real event, not an artefact of our study.</p>
<figure class="align-center zoomable">
<a href="https://images.theconversation.com/files/528832/original/file-20230529-25-wonef0.png?ixlib=rb-1.1.0&q=45&auto=format&w=1000&fit=clip"><img alt="" src="https://images.theconversation.com/files/528832/original/file-20230529-25-wonef0.png?ixlib=rb-1.1.0&q=45&auto=format&w=754&fit=clip" srcset="https://images.theconversation.com/files/528832/original/file-20230529-25-wonef0.png?ixlib=rb-1.1.0&q=45&auto=format&w=600&h=450&fit=crop&dpr=1 600w, https://images.theconversation.com/files/528832/original/file-20230529-25-wonef0.png?ixlib=rb-1.1.0&q=30&auto=format&w=600&h=450&fit=crop&dpr=2 1200w, https://images.theconversation.com/files/528832/original/file-20230529-25-wonef0.png?ixlib=rb-1.1.0&q=15&auto=format&w=600&h=450&fit=crop&dpr=3 1800w, https://images.theconversation.com/files/528832/original/file-20230529-25-wonef0.png?ixlib=rb-1.1.0&q=45&auto=format&w=754&h=566&fit=crop&dpr=1 754w, https://images.theconversation.com/files/528832/original/file-20230529-25-wonef0.png?ixlib=rb-1.1.0&q=30&auto=format&w=754&h=566&fit=crop&dpr=2 1508w, https://images.theconversation.com/files/528832/original/file-20230529-25-wonef0.png?ixlib=rb-1.1.0&q=15&auto=format&w=754&h=566&fit=crop&dpr=3 2262w" sizes="(min-width: 1466px) 754px, (max-width: 599px) 100vw, (min-width: 600px) 600px, 237px"></a>
<figcaption>
<span class="caption">Cherry blossom photographs from Flickr taken within Japan from 2008 to 2018 show an April peak as well as an unexpected smaller peak in November.</span>
<span class="attribution"><span class="source">ElQadi et al.</span>, <span class="license">Author provided</span></span>
</figcaption>
</figure>
<p>So, even without knowing it, many of us are already helping to understand how climate change influences our environment, simply by posting online photographs we capture. Dedicated sites like <a href="https://wildpollinatorcount.com/">Wild Pollinator Count</a> are excellent resources to contribute to the growing knowledge base. </p>
<p>The complex issues of climate change are still being mapped. Citizen science allows our daily observations to improve our understanding, and so better manage our relationship with the natural world.</p><img src="https://counter.theconversation.com/content/206574/count.gif" alt="The Conversation" width="1" height="1" />
<p class="fine-print"><em><span>Adrian Dyer receives funding from The Australian Research Council, and the Alexander von Humboldt Foundation (Germany).</span></em></p><p class="fine-print"><em><span>Alan Dorin receives funding and/or support from the Australian Research Council, AgriFutures, Costa Group, Australian Blueberry Growers Association and Sunny Ridge berries.</span></em></p><p class="fine-print"><em><span>Carolyn Vlasveld was undertaking a PhD at Monash University while collaborating on the study mentioned in this article. Her PhD was financially supported by an Australian Government Research Training Program (RTP) Scholarship, a Monash University Graduate Research Completion Award, an Ecological Society of Australia (ESA) Student Scholarship, a Denis and Maisie Carr Award and Travel Grant, and an Australian Society of Plant Scientists RN Robertson Travelling Fellowship.</span></em></p><p class="fine-print"><em><span>Moataz ElQadi worked on this project as part of his PhD where he was supported by an Australian Government Research Training Program (RTP) scholarship and a Monash university stipend. Moataz also received an AI for Earth grant from Microsoft.</span></em></p>Publicly available data on social media opens a new avenue for studying the environment with “incidental citizen science”.Adrian Dyer, Associate Professor, Monash UniversityAlan Dorin, Associate Professor, Faculty of Information Technology, Monash UniversityCarolyn Vlasveld, PhD candidate, Monash UniversityMoataz ElQadi, Adjunct Researcher, Faculty of Information Technology, Monash UniversityLicensed as Creative Commons – attribution, no derivatives.tag:theconversation.com,2011:article/1996882023-04-05T15:45:25Z2023-04-05T15:45:25ZAstronomers used machine learning to mine data from South Africa’s MeerKAT telescope: what they found<figure><img src="https://images.theconversation.com/files/515439/original/file-20230315-28-t0q61o.jpg?ixlib=rb-1.1.0&rect=94%2C111%2C715%2C558&q=45&auto=format&w=496&fit=clip" /><figcaption><span class="caption">SAURON: radio intensity (purple) from MeerKAT overlaid on an optical image from the Dark Energy Survey.</span> <span class="attribution"><span class="source">Michelle Lochner / The Dark Energy Survey Collaboration 2005</span></span></figcaption></figure><p>New telescopes with unprecedented sensitivity and resolution are being unveiled around the world – and beyond. Among them are the <a href="https://giantmagellan.org/">Giant Magellan Telescope</a> under construction in Chile, and the <a href="https://webb.nasa.gov/">James Webb Space Telescope</a>, which is parked a million and a half kilometres out in space. </p>
<p>This means there is a wealth of data available to scientists that simply wasn’t there before. The raw data off just a single observation from the <a href="https://www.sarao.ac.za/science/meerkat/">MeerKAT radio telescope</a> in South Africa’s Northern Cape province can measure a terabyte. That’s enough to fill a laptop computer’s hard drive. <a href="https://theconversation.com/a-big-moment-for-africa-why-the-meerkat-and-astronomy-matter-99714">MeerKAT</a> is an array of 64 large antenna dishes. It uses radio signals from space to study the evolution of the universe and everything it contains – galaxies, for example. Each dish is said to generate as much <a href="https://www.sarao.ac.za/science/meerkat/about-meerkat/">data in one second</a> as you’d find on a DVD.</p>
<p><a href="https://www.britannica.com/technology/machine-learning">Machine learning</a> is helping astronomers to work through this data quickly and more accurately than poring over it manually. Perhaps surprisingly, despite increasing reliance on computers, up until recently the discovery of rare or new astrophysical phenomena has completely relied on human inspection of the data. </p>
<p>Machine learning is essentially a set of algorithms designed to automatically learn patterns and models from data. Because we astronomers aren’t sure what we’re going to find – we don’t know what we don’t know – we also design algorithms to look out for anomalies that don’t fit known parameters or “labels”.</p>
<p>This approach allowed my colleagues and I <a href="https://academic.oup.com/mnras/advance-article-abstract/doi/10.1093/mnras/stad074/6985618?redirectedFrom=fulltext">to spot</a> a previously overlooked object in data from MeerKAT. It sits some seven billion light years from Earth (a light year is a measure of how far light would travel in a year). From what we know of the object so far, it has many of the makings of what’s known as an Odd Radio Circle (ORC). </p>
<p>Odd Radio Circles are identifiable by their <a href="https://astronomy.com/news/2022/05/understanding-the-origins-of-orcs-odd-radio-circles">strange, ring-like structure</a>. Only a handful of these circles have been detected since the first discovery in 2019, so not much is known about them yet.</p>
<hr>
<p>
<em>
<strong>
Read more:
<a href="https://theconversation.com/combined-power-of-two-telescopes-is-helping-crack-the-mystery-of-eerie-rings-in-the-sky-180595">Combined power of two telescopes is helping crack the mystery of eerie rings in the sky</a>
</strong>
</em>
</p>
<hr>
<p>In a new <a href="https://academic.oup.com/mnras/advance-article-abstract/doi/10.1093/mnras/stad074/6985618?redirectedFrom=fulltext">paper</a> we outline the features of our potential Odd Radio Circle, which we’ve named SAURON (a Steep and Uneven Ring Of Non-thermal Radiation). SAURON is, to our knowledge, the first scientific discovery made in MeerKAT data with machine learning. (There have been a handful of other discoveries assisted by machine learning in astronomy.)</p>
<p>Not only is discovering something new incredibly exciting, new discoveries are critical for challenging our understanding of the <a href="https://www.britannica.com/science/Cosmos-astronomy">cosmos</a>. These new objects may match our theories of how galaxies form and evolve, or we may need to change how we see the universe. New discoveries of anomalous astrophysical objects help science to make progress. </p>
<h2>Identifying anomalies</h2>
<p>We spotted SAURON in data from the <a href="https://arxiv.org/abs/2111.05673">MeerKAT Galaxy Cluster Legacy Survey</a>. The survey is a programme of observations conducted with South Africa’s MeerKAT telescope, a precursor to the <a href="https://www.skao.int/">Square Kilometre Array</a>. The array is a global project to build the world’s largest and most sensitive radio telescope within the coming decade, co-located in South Africa and Australia. </p>
<p>The survey was conducted between June 2018 and June 2019. It zeroed in on some 115 galaxy clusters, each made up of hundreds or even thousands of galaxies.</p>
<p>That’s a lot of data to sift through – which is where machine learning comes in. </p>
<p>We developed and used a coding framework which we called <a href="https://arxiv.org/abs/2010.11202">Astronomaly</a> to sort through the data. Astronomaly ranked unknown objects according to an anomaly scoring system. The human team then manually evaluated the 200 anomalies that interested us most. Here, we drew on vast collective expertise to make sense of the data. </p>
<p>It was during this part of the process that we identified SAURON. Instead of having to look at 6,000 individual images, we only had to look through the first 60 that Astronomaly flagged as anomalous to pick up SAURON. </p>
<p>But the question remains: what, exactly, have we found?</p>
<h2>Is SAURON an Odd Radio Circle?</h2>
<p>We know very little about Odd Radio Circles. It is currently thought that their bright, blast-like emission is the wreckage of a huge <a href="https://theconversation.com/odd-radio-circles-that-baffled-astronomers-are-likely-explosions-from-distant-galaxies-178290">explosion</a> in their host galaxies.</p>
<figure class="align-right zoomable">
<a href="https://images.theconversation.com/files/517713/original/file-20230327-23-vb272r.jpg?ixlib=rb-1.1.0&q=45&auto=format&w=1000&fit=clip"><img alt="Purple roughly circular shape on dark background" src="https://images.theconversation.com/files/517713/original/file-20230327-23-vb272r.jpg?ixlib=rb-1.1.0&q=45&auto=format&w=237&fit=clip" srcset="https://images.theconversation.com/files/517713/original/file-20230327-23-vb272r.jpg?ixlib=rb-1.1.0&q=45&auto=format&w=600&h=446&fit=crop&dpr=1 600w, https://images.theconversation.com/files/517713/original/file-20230327-23-vb272r.jpg?ixlib=rb-1.1.0&q=30&auto=format&w=600&h=446&fit=crop&dpr=2 1200w, https://images.theconversation.com/files/517713/original/file-20230327-23-vb272r.jpg?ixlib=rb-1.1.0&q=15&auto=format&w=600&h=446&fit=crop&dpr=3 1800w, https://images.theconversation.com/files/517713/original/file-20230327-23-vb272r.jpg?ixlib=rb-1.1.0&q=45&auto=format&w=754&h=561&fit=crop&dpr=1 754w, https://images.theconversation.com/files/517713/original/file-20230327-23-vb272r.jpg?ixlib=rb-1.1.0&q=30&auto=format&w=754&h=561&fit=crop&dpr=2 1508w, https://images.theconversation.com/files/517713/original/file-20230327-23-vb272r.jpg?ixlib=rb-1.1.0&q=15&auto=format&w=754&h=561&fit=crop&dpr=3 2262w" sizes="(min-width: 1466px) 754px, (max-width: 599px) 100vw, (min-width: 600px) 600px, 237px"></a>
<figcaption>
<span class="caption">SAURON.</span>
<span class="attribution"><span class="source">Michelle Lochner</span></span>
</figcaption>
</figure>
<p>The name SAURON captures the fundamentals of the object’s make-up. “Steep” refers to its spectral slope, indicating that at higher radio frequencies the “source” (or object) very quickly grows fainter. “Ring” refers to the shape. And the “Non-Thermal Radiation” refers to the type of radiation, suggesting that there must be particles accelerating in powerful magnetic fields. SAURON is at least 1.2 million light years across, about 20 times the size of the Milky Way.</p>
<p>But SAURON doesn’t tick all the right boxes for us to say that it’s definitely an Odd Radio Circle. We detected a host galaxy but can find no evidence of radio emissions with the wavelengths and frequency that match those of host galaxies of the other known ORCs. </p>
<p>And even though SAURON has a number of features in common with Odd Radio Circle1 – the first Odd Radio Circle spotted – it differs in others. Its strange shape and its oddly behaving magnetic fields don’t align well with the main structure.</p>
<p>One of the most exciting possibilities is that SAURON is a remnant of the explosive merger of two supermassive black holes. These are incredibly dense objects at the centre of galaxies such as our Milky Way that could cause a massive explosion when galaxies collide. </p>
<hr>
<p>
<em>
<strong>
Read more:
<a href="https://theconversation.com/how-were-probing-the-secrets-of-a-giant-black-hole-at-our-galaxys-centre-108181">How we're probing the secrets of a giant black hole at our galaxy's centre</a>
</strong>
</em>
</p>
<hr>
<h2>More to come</h2>
<p>More investigation is required to unravel the mystery. Meanwhile, machine learning is quickly becoming an indispensable tool to find more strange objects by sorting through enormous datasets from telescopes. With this tool, we can expect to unveil more of what the universe is hiding.</p><img src="https://counter.theconversation.com/content/199688/count.gif" alt="The Conversation" width="1" height="1" />
<p class="fine-print"><em><span>Michelle Lochner receives funding from the National Research Foundation and the Department of Science and Innovation. </span></em></p>Machine learning is becoming an indispensable tool in astronomy by sorting through enormous datasets from telescopes.Michelle Lochner, Staff Scientist at the South African Radio Astronomy Observatory and Senior Lecturer in Astronomy, University of the Western CapeLicensed as Creative Commons – attribution, no derivatives.tag:theconversation.com,2011:article/1986942023-02-01T12:08:35Z2023-02-01T12:08:35ZSeti: alien hunters get a boost as AI helps identify promising signals from space<figure><img src="https://images.theconversation.com/files/507097/original/file-20230130-12-qfen8v.jpg?ixlib=rb-1.1.0&rect=672%2C272%2C2956%2C1999&q=45&auto=format&w=496&fit=clip" /><figcaption><span class="caption">The new study analysed data gathered at the Green Bank Observatory in West Virginia.</span> <span class="attribution"><a class="source" href="https://www.shutterstock.com/image-photo/green-bank-west-virginia-october-15-762059119">Shutterstock</a></span></figcaption></figure><p>An international team of researchers looking for signs of intelligent life in space have used artificial intelligence (AI) to reveal eight promising radio signals in data collected at a US observatory.</p>
<p>The results of their research, <a href="https://www.nature.com/articles/s41550-022-01872-z">published in Nature Astronomy</a> are remarkable. The team hasn’t yet carried out an exhaustive analysis, but the paper suggests the signals have many of the characteristics we would expect if they were artificially generated. In other words, they are the kinds of signals we might pick up from an extraterrestrial civilisation broadcasting into space.</p>
<p>A cursory review of the new paper suggest these are indeed promising signals. They’re much more compelling than what is perhaps the most famous Seti candidate, <a href="https://astronomy.com/news/2020/09/the-wow-signal-an-alien-missed-connectio">the “Wow!” signal</a>, radio emission bearing the hallmarks of an extraterrestrial origin that was collected by an Ohio telescope in 1977.</p>
<p>Realistically, it’s most likely that these eight new signals were generated by human technology. But the real story here is the effectiveness of AI and <a href="https://en.wikipedia.org/wiki/Deep_learning">the techniques used by the team to</a> dig out rare and interesting signals previously buried in the noise of human-generated <a href="https://public.nrao.edu/telescopes/radio-frequency-interference/">radio frequency interference,</a> such as mobile phones and GPS.</p>
<p>Astronomers working in the field of <a href="https://www.seti.org/primer-seti-seti-institute">Seti (the search for extraterrestrial intelligence)</a> must filter out interference produced by radio communications here on Earth.</p>
<p>In this case, Peter Ma from the University of Toronto and his colleagues unleashed a set of algorithms on a mountain of data collected by the <a href="https://greenbankobservatory.org">Green Bank Telescope in West Virginia</a>, US. The data was gathered through a Seti initiative called <a href="https://seti.berkeley.edu/listen/">Breakthrough Listen</a>, established in 2015 by the investor Yuri Milner and his wife Julia. </p>
<p>Here are the characteristics astronomers look for in signals that could be artificially-generated: firstly they are <a href="https://en.wikipedia.org/wiki/Narrowband">narrow-band</a>, which means that where the radio transmission is confined to only a few frequency channels. They also disappear as the telescope is moved to another direction in the sky, and they exhibit <a href="https://en.wikipedia.org/wiki/Doppler_effect">“Doppler drifting”</a>, where the frequency of the signal changes in a predictable way with time. We would expect Doppler drifting because both the transmitter — on a distant planet, for example — and the receiver, on Earth, are moving.</p>
<figure class="align-center ">
<img alt="Artist's impression of exoplanets" src="https://images.theconversation.com/files/507060/original/file-20230130-22-kadncw.jpg?ixlib=rb-1.1.0&q=45&auto=format&w=754&fit=clip" srcset="https://images.theconversation.com/files/507060/original/file-20230130-22-kadncw.jpg?ixlib=rb-1.1.0&q=45&auto=format&w=600&h=449&fit=crop&dpr=1 600w, https://images.theconversation.com/files/507060/original/file-20230130-22-kadncw.jpg?ixlib=rb-1.1.0&q=30&auto=format&w=600&h=449&fit=crop&dpr=2 1200w, https://images.theconversation.com/files/507060/original/file-20230130-22-kadncw.jpg?ixlib=rb-1.1.0&q=15&auto=format&w=600&h=449&fit=crop&dpr=3 1800w, https://images.theconversation.com/files/507060/original/file-20230130-22-kadncw.jpg?ixlib=rb-1.1.0&q=45&auto=format&w=754&h=565&fit=crop&dpr=1 754w, https://images.theconversation.com/files/507060/original/file-20230130-22-kadncw.jpg?ixlib=rb-1.1.0&q=30&auto=format&w=754&h=565&fit=crop&dpr=2 1508w, https://images.theconversation.com/files/507060/original/file-20230130-22-kadncw.jpg?ixlib=rb-1.1.0&q=15&auto=format&w=754&h=565&fit=crop&dpr=3 2262w" sizes="(min-width: 1466px) 754px, (max-width: 599px) 100vw, (min-width: 600px) 600px, 237px">
<figcaption>
<span class="caption">Any artificial signals from deep space need to be distinguished from radio interference here on Earth.</span>
<span class="attribution"><a class="source" href="https://www.shutterstock.com/image-photo/planets-deep-space-cosmos-nebula-stars-2057080619">Shutterstock</a></span>
</figcaption>
</figure>
<h2>Buried in the noise</h2>
<p>The Breakthrough Listen project’s <a href="https://seti.berkeley.edu/blc1/">first candidate signal</a>, called BLC1, was first announced in 2020. But it was <a href="https://www.nature.com/articles/s41550-021-01508-8">later traced</a> to transmissions associated with cheap electronic devices on this planet. The application of AI techniques to the Breakthrough Listen observing programme, however, is a potential game changer for the field. Even seasoned Seti researchers are beginning to think that we might be on the cusp of a momentous scientific breakthrough.</p>
<p>This may explain renewed interest by groups around the world that are planning for Seti success. For example, a <a href="https://seti.wp.st-andrews.ac.uk">Seti post-detection hub</a> has been set up at the University of St Andrews in Scotland. This will study how humans should react if we discover we are not alone in the Universe.</p>
<p>The International Academy of Astronautics (IAA) <a href="https://iaaseti.org/en/">Seti permanent committee</a> oversees the <a href="https://iaaseti.org/en/protocols/">Seti post-detection protocols</a>, which outline what steps scientists should take in the event of detecting a genuine signal. The IAA has opted to update the text of the protocols sometime later this year.</p>
<p>But the new study highlights a problem with previous signals of interest. When the team took another look at the stars associated with the eight narrow-band transmissions, they could no longer detect the signals. </p>
<p>It would not be surprising if many, and perhaps the vast majority of bona-fide Seti signals, were isolated events. After all, what are the chances that we point our telescopes in exactly the right direction, at the right time and with the right frequency on multiple occasions?</p>
<h2>Missing ingredients</h2>
<p>As I <a href="https://theconversation.com/seti-new-signal-excites-alien-hunters-heres-how-we-could-find-out-if-its-real-152498">argued here</a> a few years ago, Seti surveys would greatly benefit from employing multiple radio telescopes, operating in a manner that’s known as a <a href="https://public.nrao.edu/ask/how-does-a-radio-interferometer-work/">classical interferometer network</a>. </p>
<p>These telescope arrays (groups of several antennas observing together) generate huge amounts of data. With AI onboard, the challenge is perhaps more manageable than previously thought. </p>
<p>Breakthrough Listen is already using telescope arrays such as <a href="https://www.sarao.ac.za/science/meerkat/about-meerkat/">MeerKAT in South Africa</a> for Seti searches. In Europe, researchers have been experimenting with <a href="https://www.evlbi.org">arrays that span the globe</a>.</p>
<p>This European approach would help us isolate signals from human-made interference, give us multiple independent detections of individual events, and permit us to localise signals to individual stars and possibly orbiting planets. </p>
<p>Among the future projects is the <a href="https://www.skao.int/en">Square Kilometre Array</a>, an international project to build the two largest telescope arrays in the world, which will be based in Australia and South Africa. Another upcoming project is the <a href="https://ngvla.nrao.edu">next generation VLA (ngVLA)</a>, a series of linked telescope facilities that will be spread across the United States. These radio telescope arrays will be even more sensitive than current instruments.</p>
<p>It’s my belief — and indeed hope — that somewhere out there intelligent beings are waiting to be discovered. The AI revolution might be the missing ingredient that previous endeavours have lacked. In particular, AI algorithms will eventually evolve into powerful tools that no longer suffer from <a href="https://www.nist.gov/news-events/news/2022/03/theres-more-ai-bias-biased-data-nist-report-highlights">human biases</a>. </p>
<p>Lord Martin Rees, chairman of the Breakthrough Listen advisory board and the astronomer royal, has proposed that if we do find aliens they are likely to be <a href="https://theconversation.com/seti-why-extraterrestrial-intelligence-is-more-likely-to-be-artificial-than-biological-169966">intelligent machines</a> operating in the depths of space, unconstrained by the biological limitations placed on humans. </p>
<p>If we ever do find a bona-fide signal, it could just be that it’s mediated by machines on Earth and in space.</p><img src="https://counter.theconversation.com/content/198694/count.gif" alt="The Conversation" width="1" height="1" />
<p class="fine-print"><em><span>Michael Garrett is on the advisory board of the Breakthrough Listen initiative and the Seti Institute.</span></em></p>Can artificial intelligence transform the search for alien intelligence?Michael Garrett, Sir Bernard Lovell chair of Astrophysics and Director of Jodrell Bank Centre for Astrophysics, University of ManchesterLicensed as Creative Commons – attribution, no derivatives.tag:theconversation.com,2011:article/1962222022-12-14T19:03:03Z2022-12-14T19:03:03ZSpotting plastic waste from space and counting the fish in the seas: here’s how AI can help protect the oceans<figure><img src="https://images.theconversation.com/files/500920/original/file-20221214-3038-lf1qbe.jpg?ixlib=rb-1.1.0&rect=26%2C40%2C2968%2C1971&q=45&auto=format&w=496&fit=clip" /><figcaption><span class="caption">
</span> <span class="attribution"><span class="source">Shutterstock</span></span></figcaption></figure><p>You’ve seen the art AI image generators can create, and you may have played with natural language AI chatbots. You’ve benefited from artificial intelligence tools recommending you music and suggesting your next streaming show. </p>
<p>But AI can do much more. Humans are excellent at spotting patterns. It’s why we see <a href="https://science.nasa.gov/science-news/science-at-nasa/2001/ast24may_1">faces on Mars</a> or in the clouds. But in some areas, AI is even better. Give one of these tools a million photographs and ask it to spot telltale signs – and it can. AI can enable research at scales previously impossible. </p>
<p>We’ve used AI’s exceptional pattern recognition to trawl through satellite images and map the tonnes of plastic pollution threatening our seas – in real time. Already, this technique <a href="https://www.minderoo.org/global-plastic-watch/">has found</a> more than 4,000 unreported informal dumps next to rivers. This is useful, given <a href="https://www.scientificamerican.com/article/stemming-the-plastic-tide-10-rivers-contribute-most-of-the-plastic-in-the-oceans/">just ten rivers</a> contribute nearly all the plastic entering our oceans. </p>
<p>This is just the start. So far, AI has shown promise in <a href="https://www.minderoo.org/flourishing-oceans/">our projects</a> mapping seagrass meadows from space and finding unknown reefs likely to harbour heat-resilient coral. Soon, we hope we’ll be able to put AI on the job to find out exactly what fish live where – without ever seeing them. </p>
<figure class="align-center zoomable">
<a href="https://images.theconversation.com/files/500921/original/file-20221214-3038-9c9xua.jpg?ixlib=rb-1.1.0&q=45&auto=format&w=1000&fit=clip"><img alt="fish school reef" src="https://images.theconversation.com/files/500921/original/file-20221214-3038-9c9xua.jpg?ixlib=rb-1.1.0&q=45&auto=format&w=754&fit=clip" srcset="https://images.theconversation.com/files/500921/original/file-20221214-3038-9c9xua.jpg?ixlib=rb-1.1.0&q=45&auto=format&w=600&h=338&fit=crop&dpr=1 600w, https://images.theconversation.com/files/500921/original/file-20221214-3038-9c9xua.jpg?ixlib=rb-1.1.0&q=30&auto=format&w=600&h=338&fit=crop&dpr=2 1200w, https://images.theconversation.com/files/500921/original/file-20221214-3038-9c9xua.jpg?ixlib=rb-1.1.0&q=15&auto=format&w=600&h=338&fit=crop&dpr=3 1800w, https://images.theconversation.com/files/500921/original/file-20221214-3038-9c9xua.jpg?ixlib=rb-1.1.0&q=45&auto=format&w=754&h=424&fit=crop&dpr=1 754w, https://images.theconversation.com/files/500921/original/file-20221214-3038-9c9xua.jpg?ixlib=rb-1.1.0&q=30&auto=format&w=754&h=424&fit=crop&dpr=2 1508w, https://images.theconversation.com/files/500921/original/file-20221214-3038-9c9xua.jpg?ixlib=rb-1.1.0&q=15&auto=format&w=754&h=424&fit=crop&dpr=3 2262w" sizes="(min-width: 1466px) 754px, (max-width: 599px) 100vw, (min-width: 600px) 600px, 237px"></a>
<figcaption>
<span class="caption">How many fish are there in the sea? AI could make possible real-time tracking of fish species.</span>
<span class="attribution"><span class="source">Shutterstock</span></span>
</figcaption>
</figure>
<h2>Is AI really a gamechanger for science?</h2>
<p>Yes. Think of the vast volumes of data scientists have gathered in recent decades. Until now, trawling through the data has been painstaking and at times tedious. That’s because while detecting patterns is something humans do well, we’re slower. </p>
<p>AI mines large data sets, which can be anything from photos to numbers. You train it so it knows what you’re looking for. Then the software tool gets to work, detecting patterns – and importantly, offering up predictions about how these patterns arise. </p>
<hr>
<p>
<em>
<strong>
Read more:
<a href="https://theconversation.com/how-shoring-up-drones-with-artificial-intelligence-helps-surf-lifesavers-spot-sharks-at-the-beach-192498">How shoring up drones with artificial intelligence helps surf lifesavers spot sharks at the beach</a>
</strong>
</em>
</p>
<hr>
<p>These methods are especially powerful for messy and complex biological data. For example, the AI tool AlphaFold has <a href="https://cosmosmagazine.com/technology/googles-protein-folding-ai-alphafold/">totally revolutionised</a> the slow process of understanding how proteins fold themselves into origami-like shapes inside cells. Previously, it might have taken months or years to figure out a single protein structure. This year, AlphaFold announced predicted structures for <a href="https://alphafold.ebi.ac.uk/">200 million proteins</a>. </p>
<h2>What can AI offer ecology?</h2>
<p>We’ve found AI useful at <a href="https://www.nature.com/articles/s41467-022-28956-8">finding unknown reefs</a> with corals primed to survive despite warming waters. That’s vital, given the oceans have taken up almost all of the heat trapped by the trillion tonnes of greenhouse gases we’ve put in the atmosphere. </p>
<p>And we’ve found AI can usefully identify specific environmental conditions under which reefs will survive as the oceans heat up. Our research suggests hundreds of reefs among the thousands in the Great Barrier Reef <a href="https://www.nature.com/articles/s41467-022-28956-8">may be home</a> to corals which have higher heat tolerance than normal. Now we know this, we can protect these reefs – and turn to them for potential use to restore dying reefs elsewhere. </p>
<figure class="align-center zoomable">
<a href="https://images.theconversation.com/files/500926/original/file-20221214-24-i1dktw.jpg?ixlib=rb-1.1.0&q=45&auto=format&w=1000&fit=clip"><img alt="coral fragment" src="https://images.theconversation.com/files/500926/original/file-20221214-24-i1dktw.jpg?ixlib=rb-1.1.0&q=45&auto=format&w=754&fit=clip" srcset="https://images.theconversation.com/files/500926/original/file-20221214-24-i1dktw.jpg?ixlib=rb-1.1.0&q=45&auto=format&w=600&h=450&fit=crop&dpr=1 600w, https://images.theconversation.com/files/500926/original/file-20221214-24-i1dktw.jpg?ixlib=rb-1.1.0&q=30&auto=format&w=600&h=450&fit=crop&dpr=2 1200w, https://images.theconversation.com/files/500926/original/file-20221214-24-i1dktw.jpg?ixlib=rb-1.1.0&q=15&auto=format&w=600&h=450&fit=crop&dpr=3 1800w, https://images.theconversation.com/files/500926/original/file-20221214-24-i1dktw.jpg?ixlib=rb-1.1.0&q=45&auto=format&w=754&h=566&fit=crop&dpr=1 754w, https://images.theconversation.com/files/500926/original/file-20221214-24-i1dktw.jpg?ixlib=rb-1.1.0&q=30&auto=format&w=754&h=566&fit=crop&dpr=2 1508w, https://images.theconversation.com/files/500926/original/file-20221214-24-i1dktw.jpg?ixlib=rb-1.1.0&q=15&auto=format&w=754&h=566&fit=crop&dpr=3 2262w" sizes="(min-width: 1466px) 754px, (max-width: 599px) 100vw, (min-width: 600px) 600px, 237px"></a>
<figcaption>
<span class="caption">Finding naturally heat-resilient corals could help us safeguard reef ecosystems.</span>
<span class="attribution"><span class="source">Shutterstock</span></span>
</figcaption>
</figure>
<p>This idea of “super reefs” isn’t new. Other researchers have focused on protecting <a href="https://www.50reefs.org/">50 coral reefs globally</a> in the hope of safeguarding these ecosystems against the expected mass coral death as water temperatures rise. What we have added was the discovery that AI can help find these heat-resilient corals. Without AI, it would have been like trying to find a needle in a haystack. </p>
<p>Spotting plastic waste from space would have been almost impossible before AI image detection programs became available. How does it work? Essentially, photos taken by European Space Agency satellites are scanned by AI to spot hidden plastic dumps. Then we refine it over time, to see if these sites are getting bigger – and if they’re close to rivers or lakes, which could carry plastics into the seas and add to the millions of tonnes of turtle-choking, fish-killing plastics already swilling around. </p>
<p>The goal is to find the sites at highest risk of adding to ocean plastics. Once we know this, enforcement agencies in each of the 112 countries we’ve mapped can respond to the most urgent problems first. So far <a href="https://arxiv.org/abs/2204.01485">we’ve found</a> more than 4,000 sites, with around one in five within 200 metres of a waterway. When we looked at Indonesia in detail, we found double the number of publicly listed dump sites.</p>
<figure class="align-center zoomable">
<a href="https://images.theconversation.com/files/500923/original/file-20221214-22-nds9cp.jpg?ixlib=rb-1.1.0&q=45&auto=format&w=1000&fit=clip"><img alt="Irawaddy delta from space" src="https://images.theconversation.com/files/500923/original/file-20221214-22-nds9cp.jpg?ixlib=rb-1.1.0&q=45&auto=format&w=754&fit=clip" srcset="https://images.theconversation.com/files/500923/original/file-20221214-22-nds9cp.jpg?ixlib=rb-1.1.0&q=45&auto=format&w=600&h=400&fit=crop&dpr=1 600w, https://images.theconversation.com/files/500923/original/file-20221214-22-nds9cp.jpg?ixlib=rb-1.1.0&q=30&auto=format&w=600&h=400&fit=crop&dpr=2 1200w, https://images.theconversation.com/files/500923/original/file-20221214-22-nds9cp.jpg?ixlib=rb-1.1.0&q=15&auto=format&w=600&h=400&fit=crop&dpr=3 1800w, https://images.theconversation.com/files/500923/original/file-20221214-22-nds9cp.jpg?ixlib=rb-1.1.0&q=45&auto=format&w=754&h=503&fit=crop&dpr=1 754w, https://images.theconversation.com/files/500923/original/file-20221214-22-nds9cp.jpg?ixlib=rb-1.1.0&q=30&auto=format&w=754&h=503&fit=crop&dpr=2 1508w, https://images.theconversation.com/files/500923/original/file-20221214-22-nds9cp.jpg?ixlib=rb-1.1.0&q=15&auto=format&w=754&h=503&fit=crop&dpr=3 2262w" sizes="(min-width: 1466px) 754px, (max-width: 599px) 100vw, (min-width: 600px) 600px, 237px"></a>
<figcaption>
<span class="caption">You could trawl through millions of satellite images looking for hidden plastic waste dumps - or you could train an AI tool and let the software do the work.</span>
<span class="attribution"><span class="source">ESA/Shutterstock</span></span>
</figcaption>
</figure>
<p>AI is also proving itself as a labour-saver. One part of science often hidden to the general public is the sheer number of manual, repetitive tasks. For instance, if you want to figure out why some baby coral polyps survive heat or more acid water while others die, you have to measure colour, growth and survival rates over time. We’ve found AI can <a href="https://www.mdpi.com/2072-4292/13/16/3173">do this work</a>, precisely and fast. </p>
<p>Of course, AI is not magic. It is a tool, and all tools have pitfalls. One problem is placing too much trust in AI outputs, believing them true because the algorithm has seen more data than we have. But this is dangerous, as the <a href="https://www.theverge.com/2022/12/5/23493932/chatgpt-ai-generated-answers-temporarily-banned-stack-overflow-llms-dangers">confidently wrong</a> answers given by the new ChatGPT AI demonstrate. </p>
<p>Ecology isn’t free from biases either. That means we have to carefully evaluate the data we use to train the AI. Plus, we have to remain vigilant and manually evaluate AI predictions to figure out if they fit with our reality. AI is a valuable assistant for ecologists – not a replacement.</p>
<figure class="align-center zoomable">
<a href="https://images.theconversation.com/files/500925/original/file-20221214-24-h5elwz.jpg?ixlib=rb-1.1.0&q=45&auto=format&w=1000&fit=clip"><img alt="floating weather buoy" src="https://images.theconversation.com/files/500925/original/file-20221214-24-h5elwz.jpg?ixlib=rb-1.1.0&q=45&auto=format&w=754&fit=clip" srcset="https://images.theconversation.com/files/500925/original/file-20221214-24-h5elwz.jpg?ixlib=rb-1.1.0&q=45&auto=format&w=600&h=400&fit=crop&dpr=1 600w, https://images.theconversation.com/files/500925/original/file-20221214-24-h5elwz.jpg?ixlib=rb-1.1.0&q=30&auto=format&w=600&h=400&fit=crop&dpr=2 1200w, https://images.theconversation.com/files/500925/original/file-20221214-24-h5elwz.jpg?ixlib=rb-1.1.0&q=15&auto=format&w=600&h=400&fit=crop&dpr=3 1800w, https://images.theconversation.com/files/500925/original/file-20221214-24-h5elwz.jpg?ixlib=rb-1.1.0&q=45&auto=format&w=754&h=503&fit=crop&dpr=1 754w, https://images.theconversation.com/files/500925/original/file-20221214-24-h5elwz.jpg?ixlib=rb-1.1.0&q=30&auto=format&w=754&h=503&fit=crop&dpr=2 1508w, https://images.theconversation.com/files/500925/original/file-20221214-24-h5elwz.jpg?ixlib=rb-1.1.0&q=15&auto=format&w=754&h=503&fit=crop&dpr=3 2262w" sizes="(min-width: 1466px) 754px, (max-width: 599px) 100vw, (min-width: 600px) 600px, 237px"></a>
<figcaption>
<span class="caption">Automatic weather buoys already exist. It won’t be long until we can build AI water sampling drones to tell us about life underneath.</span>
<span class="attribution"><span class="source">Shutterstock</span></span>
</figcaption>
</figure>
<h2>What’s next?</h2>
<p>Imagine having autonomous floating or underwater drones sampling seawater, with AI neural networks looking for fish DNA. It sounds like sci-fi, but it’s now entirely possible. Drone technology has matured. AI tools have arrived. And we no longer need to catch fish to know what lives in the seas. All you need are tiny traces of <a href="https://www.usgs.gov/special-topics/water-science-school/science/environmental-dna-edna">environmental DNA</a> marine species leave behind in water. Similarly, we could track coral reef ecosystem health in near real time. </p>
<p>This will let us take the pulse of these ecosystems at a time when our oceans are under unprecedented pressure from industrial fishing, marine heatwaves and acidification from climate change, and plastic pollution. The more we know, the better we can respond. </p>
<hr>
<p>
<em>
<strong>
Read more:
<a href="https://theconversation.com/ai-can-tackle-the-climate-emergency-if-developed-responsibly-132908">AI can tackle the climate emergency – if developed responsibly</a>
</strong>
</em>
</p>
<hr>
<img src="https://counter.theconversation.com/content/196222/count.gif" alt="The Conversation" width="1" height="1" />
<p class="fine-print"><em><span>Philipp Bayer receives funding from Minderoo Foundation. He is affiliated with Minderoo Foundation. </span></em></p><p class="fine-print"><em><span>Ahmed Elagali receives funding from Minderoo Foundation. He is affiliated with Minderoo Foundation</span></em></p><p class="fine-print"><em><span>Julie Robidart receives funding from the Minderoo Foundation. She is affiliated with the Minderoo Foundation.</span></em></p><p class="fine-print"><em><span>Kate Marie Quigley receives funding from the Minderoo Foundation. She is affiliated with the Minderoo Foundation.</span></em></p>Humans are expert pattern-finders. But artificial intelligence tools are better at trawling through vast data sets to find anything from waste dumps to heat-tolerant corals.Philipp Bayer, Adjunct Research Fellow, UWA Oceans Institute, The University of Western AustraliaAhmed Elagali, Research associate, The University of Western AustraliaJulie Robidart, Adjunct Senior Research Fellow, UWA Oceans Institute, The University of Western AustraliaKate Marie Quigley, Adjunct Senior Research Fellow, James Cook UniversityLicensed as Creative Commons – attribution, no derivatives.tag:theconversation.com,2011:article/1964182022-12-12T17:52:28Z2022-12-12T17:52:28ZWorld Cup 2022: crunching 150 years of big data to predict the winner<figure><img src="https://images.theconversation.com/files/500469/original/file-20221212-113843-qh64s7.jpg?ixlib=rb-1.1.0&rect=30%2C25%2C3402%2C2235&q=45&auto=format&w=496&fit=clip" /><figcaption><span class="caption">France's goalkeeper #01 Hugo Lloris (C) jumps for the ball during the Qatar 2022 World Cup quarter-final football match between England and France at the Al-Bayt Stadium in Al Khor, north of Doha, on December 10, 2022. </span> <span class="attribution"><a class="source" href="https://news.afp.com/#/c/main/search/photos?id=newsml.afp.com.20221210T210326Z.doc-33362bz&type=photo">Jewel Samad/AFP</a></span></figcaption></figure><p>Perhaps more than any recent World Cup, this year’s competition in Qatar has thrown up considerable surprises. Who indeed, of the analytics crowd, could have predicted Saudi Arabia would defeat Argentina the way it did, or competition-favourite Brazil would end up losing to Croatia? Meanwhile, Morocco has stunned commentators by becoming <a href="https://edition.cnn.com/2022/12/10/football/morocco-portugal-quarterfinals-world-cup-2022-spt-intl/index.html">the first African nation to reach semi-finals</a>. Now, almost a week away from the final, speculation is about who will win the game is at its apex. Is there any way we could predict the results better than we did by “following the science”? </p>
<h2>150 years of big data</h2>
<p>In collaboration with analytics company <a href="https://www.alteryx.com/">Alteryx</a> and Stirling University , our team at Audencia has given its best shot by developing a sophisticated World Cup prediction model drawing from 150 years of international football match results, including tournaments and friendly games. </p>
<p>While developing our initial mathematical model, we considered factors such as win ratio, goals scored, and overall match results. To further improve the accuracy of our prediction, we took into account individual teams’ current FIFA ranking and overall rating. We also added <a href="https://www.fifaratings.com/">FIFA player ratings</a> along with individual player skills and attribute scores (i.e., attack, movement, power, defence). This allows fine adjustments in our modelling technique based on individual player selection and injuries during later stages.</p>
<p>With the existing data as a <a href="https://theconversation.com/what-is-machine-learning-76759">machine-learning training mechanism</a>, we employed <a href="https://www.ibm.com/cloud/learn/random-forest">Random Forest algorithm</a> to predict results for every World Cup fixture. Using the <a href="https://www.alteryx.com/">Alteryx</a> data analysis platform, we calculated the overall outcome of individual games along with <a href="https://understat.com/">expected goals</a> (xG) per team per match. Overall, our model showed 60% to 70% accuracy rate in the course of the training and testing phase. In data science language, this is considered to be a range of accuracy that is acceptable to good in predicting the outcome of an event.</p>
<figure class="align-center ">
<img alt="" src="https://images.theconversation.com/files/500407/original/file-20221212-109624-4pofku.png?ixlib=rb-1.1.0&q=45&auto=format&w=754&fit=clip" srcset="https://images.theconversation.com/files/500407/original/file-20221212-109624-4pofku.png?ixlib=rb-1.1.0&q=45&auto=format&w=600&h=331&fit=crop&dpr=1 600w, https://images.theconversation.com/files/500407/original/file-20221212-109624-4pofku.png?ixlib=rb-1.1.0&q=30&auto=format&w=600&h=331&fit=crop&dpr=2 1200w, https://images.theconversation.com/files/500407/original/file-20221212-109624-4pofku.png?ixlib=rb-1.1.0&q=15&auto=format&w=600&h=331&fit=crop&dpr=3 1800w, https://images.theconversation.com/files/500407/original/file-20221212-109624-4pofku.png?ixlib=rb-1.1.0&q=45&auto=format&w=754&h=416&fit=crop&dpr=1 754w, https://images.theconversation.com/files/500407/original/file-20221212-109624-4pofku.png?ixlib=rb-1.1.0&q=30&auto=format&w=754&h=416&fit=crop&dpr=2 1508w, https://images.theconversation.com/files/500407/original/file-20221212-109624-4pofku.png?ixlib=rb-1.1.0&q=15&auto=format&w=754&h=416&fit=crop&dpr=3 2262w" sizes="(min-width: 1466px) 754px, (max-width: 599px) 100vw, (min-width: 600px) 600px, 237px">
<figcaption>
<span class="caption">Importance and ranking of key data variables used for match prediction.</span>
<span class="attribution"><span class="license">Author provided</span></span>
</figcaption>
</figure>
<h2>Group stage, knockout stage and the winner</h2>
<p>Applying the model to the FIFA World Cup 2022 group stage fixtures produced some interesting and unexpected results. We ran the simulation through 500 different sets of probabilities to verify the accuracy of these predictions. Our algorithm successfully predicted the qualifications of 11 teams, including Senegal and Morocco, reflecting 68.7% accuracy.</p>
<p><strong>Group stage result prediction</strong></p>
<figure class="align-center zoomable">
<a href="https://images.theconversation.com/files/500457/original/file-20221212-113720-4pofku.jpg?ixlib=rb-1.1.0&q=45&auto=format&w=1000&fit=clip"><img alt="" src="https://images.theconversation.com/files/500457/original/file-20221212-113720-4pofku.jpg?ixlib=rb-1.1.0&q=45&auto=format&w=754&fit=clip" srcset="https://images.theconversation.com/files/500457/original/file-20221212-113720-4pofku.jpg?ixlib=rb-1.1.0&q=45&auto=format&w=600&h=227&fit=crop&dpr=1 600w, https://images.theconversation.com/files/500457/original/file-20221212-113720-4pofku.jpg?ixlib=rb-1.1.0&q=30&auto=format&w=600&h=227&fit=crop&dpr=2 1200w, https://images.theconversation.com/files/500457/original/file-20221212-113720-4pofku.jpg?ixlib=rb-1.1.0&q=15&auto=format&w=600&h=227&fit=crop&dpr=3 1800w, https://images.theconversation.com/files/500457/original/file-20221212-113720-4pofku.jpg?ixlib=rb-1.1.0&q=45&auto=format&w=754&h=286&fit=crop&dpr=1 754w, https://images.theconversation.com/files/500457/original/file-20221212-113720-4pofku.jpg?ixlib=rb-1.1.0&q=30&auto=format&w=754&h=286&fit=crop&dpr=2 1508w, https://images.theconversation.com/files/500457/original/file-20221212-113720-4pofku.jpg?ixlib=rb-1.1.0&q=15&auto=format&w=754&h=286&fit=crop&dpr=3 2262w" sizes="(min-width: 1466px) 754px, (max-width: 599px) 100vw, (min-width: 600px) 600px, 237px"></a>
<figcaption>
<span class="caption"></span>
<span class="attribution"><span class="license">Author provided</span></span>
</figcaption>
</figure>
<p>After the start of the knockout stage, we reset the results and ran the simulation mimicking the original knockout fixtures. The new analysis also took into account player performance during this World Cup and likely player selection during each match. During the round of 16 predictions, our algorithm predicted seven correct match result outcomes, reflecting 87.5% accuracy. The only shock result was Morocco’s win over Spain, which we couldn’t capture appropriately.</p>
<p><strong>Round of 16 results (predicted vs actual)</strong></p>
<figure class="align-center zoomable">
<a href="https://images.theconversation.com/files/500458/original/file-20221212-105279-f6pez3.jpg?ixlib=rb-1.1.0&q=45&auto=format&w=1000&fit=clip"><img alt="" src="https://images.theconversation.com/files/500458/original/file-20221212-105279-f6pez3.jpg?ixlib=rb-1.1.0&q=45&auto=format&w=754&fit=clip" srcset="https://images.theconversation.com/files/500458/original/file-20221212-105279-f6pez3.jpg?ixlib=rb-1.1.0&q=45&auto=format&w=600&h=144&fit=crop&dpr=1 600w, https://images.theconversation.com/files/500458/original/file-20221212-105279-f6pez3.jpg?ixlib=rb-1.1.0&q=30&auto=format&w=600&h=144&fit=crop&dpr=2 1200w, https://images.theconversation.com/files/500458/original/file-20221212-105279-f6pez3.jpg?ixlib=rb-1.1.0&q=15&auto=format&w=600&h=144&fit=crop&dpr=3 1800w, https://images.theconversation.com/files/500458/original/file-20221212-105279-f6pez3.jpg?ixlib=rb-1.1.0&q=45&auto=format&w=754&h=181&fit=crop&dpr=1 754w, https://images.theconversation.com/files/500458/original/file-20221212-105279-f6pez3.jpg?ixlib=rb-1.1.0&q=30&auto=format&w=754&h=181&fit=crop&dpr=2 1508w, https://images.theconversation.com/files/500458/original/file-20221212-105279-f6pez3.jpg?ixlib=rb-1.1.0&q=15&auto=format&w=754&h=181&fit=crop&dpr=3 2262w" sizes="(min-width: 1466px) 754px, (max-width: 599px) 100vw, (min-width: 600px) 600px, 237px"></a>
<figcaption>
<span class="caption"></span>
<span class="attribution"><a class="source" href="https://www.telegraph.co.uk/world-cup/2022-predictor-wallchart-football-bracket-predictions/">Developed using The Telegraph Wallchart</a>, <span class="license">Author provided</span></span>
</figcaption>
</figure>
<p>In the simulation of the quarter finals, once again, we revised the algorithm considering quarterfinal fixtures and individual player’s performance during the World Cup. This time our algorithm only came out with 50% accuracy, failing to predict Brazil’s exit and Morocco’s triumph over Portugal. The tournament and fan’s favourite, Brazil’s loss to Croatia was a result of their failure to create early scoring chances. Morocco and Croatia’s persistent resistance leading to elimination of bigger team show that this World Cup has favoured teams who were well organised with their defence.</p>
<p>Another important aspect of this world cup is penalty shootout success. Big teams like Spain and Brazil fell due to poor penalty performance.</p>
<p><strong>Quarter-final results (predicted vs actual)</strong></p>
<figure class="align-center zoomable">
<a href="https://images.theconversation.com/files/500459/original/file-20221212-110120-k8glwk.jpg?ixlib=rb-1.1.0&q=45&auto=format&w=1000&fit=clip"><img alt="" src="https://images.theconversation.com/files/500459/original/file-20221212-110120-k8glwk.jpg?ixlib=rb-1.1.0&q=45&auto=format&w=754&fit=clip" srcset="https://images.theconversation.com/files/500459/original/file-20221212-110120-k8glwk.jpg?ixlib=rb-1.1.0&q=45&auto=format&w=600&h=69&fit=crop&dpr=1 600w, https://images.theconversation.com/files/500459/original/file-20221212-110120-k8glwk.jpg?ixlib=rb-1.1.0&q=30&auto=format&w=600&h=69&fit=crop&dpr=2 1200w, https://images.theconversation.com/files/500459/original/file-20221212-110120-k8glwk.jpg?ixlib=rb-1.1.0&q=15&auto=format&w=600&h=69&fit=crop&dpr=3 1800w, https://images.theconversation.com/files/500459/original/file-20221212-110120-k8glwk.jpg?ixlib=rb-1.1.0&q=45&auto=format&w=754&h=87&fit=crop&dpr=1 754w, https://images.theconversation.com/files/500459/original/file-20221212-110120-k8glwk.jpg?ixlib=rb-1.1.0&q=30&auto=format&w=754&h=87&fit=crop&dpr=2 1508w, https://images.theconversation.com/files/500459/original/file-20221212-110120-k8glwk.jpg?ixlib=rb-1.1.0&q=15&auto=format&w=754&h=87&fit=crop&dpr=3 2262w" sizes="(min-width: 1466px) 754px, (max-width: 599px) 100vw, (min-width: 600px) 600px, 237px"></a>
<figcaption>
<span class="caption"></span>
<span class="attribution"><a class="source" href="https://www.telegraph.co.uk/world-cup/2022-predictor-wallchart-football-bracket-predictions/">Developed using The Telegraph Wallchart</a>, <span class="license">Author provided</span></span>
</figcaption>
</figure>
<p>Similarly, England fell short as Harry Kane lost a second decisive penalty kick against France. Looking at these emerging statistics, it seems teams that can keep their goalkeepers in top forms and have a good penalty squad are likely to win the World Cup. </p>
<p>Despite having the best shot saving goalkeeper, Argentina is likely to be at disadvantage in the semi-finals due to two key player suspensions, <a href="https://www.football.london/international-football/argentina-netherlands-fifa-messi-paredes-25722319">Marcos Acuna and Gonzalo Montiel</a>. The suspension of <a href="https://www.football.london/international-football/world-cup-yellow-card-rules-25702167">Walid Cheddira</a> will also put Morocco at a disadvantage against France.</p>
<p><strong>Semi-final results (predicted)</strong></p>
<figure class="align-center zoomable">
<a href="https://images.theconversation.com/files/500459/original/file-20221212-110120-k8glwk.jpg?ixlib=rb-1.1.0&q=45&auto=format&w=1000&fit=clip"><img alt="" src="https://images.theconversation.com/files/500459/original/file-20221212-110120-k8glwk.jpg?ixlib=rb-1.1.0&q=45&auto=format&w=754&fit=clip" srcset="https://images.theconversation.com/files/500459/original/file-20221212-110120-k8glwk.jpg?ixlib=rb-1.1.0&q=45&auto=format&w=600&h=69&fit=crop&dpr=1 600w, https://images.theconversation.com/files/500459/original/file-20221212-110120-k8glwk.jpg?ixlib=rb-1.1.0&q=30&auto=format&w=600&h=69&fit=crop&dpr=2 1200w, https://images.theconversation.com/files/500459/original/file-20221212-110120-k8glwk.jpg?ixlib=rb-1.1.0&q=15&auto=format&w=600&h=69&fit=crop&dpr=3 1800w, https://images.theconversation.com/files/500459/original/file-20221212-110120-k8glwk.jpg?ixlib=rb-1.1.0&q=45&auto=format&w=754&h=87&fit=crop&dpr=1 754w, https://images.theconversation.com/files/500459/original/file-20221212-110120-k8glwk.jpg?ixlib=rb-1.1.0&q=30&auto=format&w=754&h=87&fit=crop&dpr=2 1508w, https://images.theconversation.com/files/500459/original/file-20221212-110120-k8glwk.jpg?ixlib=rb-1.1.0&q=15&auto=format&w=754&h=87&fit=crop&dpr=3 2262w" sizes="(min-width: 1466px) 754px, (max-width: 599px) 100vw, (min-width: 600px) 600px, 237px"></a>
<figcaption>
<span class="caption"></span>
<span class="attribution"><a class="source" href="https://www.telegraph.co.uk/world-cup/2022-predictor-wallchart-football-bracket-predictions/">Developed using The Telegraph Wallchart</a>, <span class="license">Author provided</span></span>
</figcaption>
</figure>
<p>Both France and Argentina have boasted tough opposition with excellent defence and goal conceding records in this World Cup. If the semi-final games end in a penalty shootout then Croatia and Morocco will have a greater chance to reach the final.</p>
<figure class="align-center zoomable">
<a href="https://images.theconversation.com/files/500461/original/file-20221212-110235-xogfk5.jpg?ixlib=rb-1.1.0&q=45&auto=format&w=1000&fit=clip"><img alt="" src="https://images.theconversation.com/files/500461/original/file-20221212-110235-xogfk5.jpg?ixlib=rb-1.1.0&q=45&auto=format&w=754&fit=clip" srcset="https://images.theconversation.com/files/500461/original/file-20221212-110235-xogfk5.jpg?ixlib=rb-1.1.0&q=45&auto=format&w=600&h=112&fit=crop&dpr=1 600w, https://images.theconversation.com/files/500461/original/file-20221212-110235-xogfk5.jpg?ixlib=rb-1.1.0&q=30&auto=format&w=600&h=112&fit=crop&dpr=2 1200w, https://images.theconversation.com/files/500461/original/file-20221212-110235-xogfk5.jpg?ixlib=rb-1.1.0&q=15&auto=format&w=600&h=112&fit=crop&dpr=3 1800w, https://images.theconversation.com/files/500461/original/file-20221212-110235-xogfk5.jpg?ixlib=rb-1.1.0&q=45&auto=format&w=754&h=141&fit=crop&dpr=1 754w, https://images.theconversation.com/files/500461/original/file-20221212-110235-xogfk5.jpg?ixlib=rb-1.1.0&q=30&auto=format&w=754&h=141&fit=crop&dpr=2 1508w, https://images.theconversation.com/files/500461/original/file-20221212-110235-xogfk5.jpg?ixlib=rb-1.1.0&q=15&auto=format&w=754&h=141&fit=crop&dpr=3 2262w" sizes="(min-width: 1466px) 754px, (max-width: 599px) 100vw, (min-width: 600px) 600px, 237px"></a>
<figcaption>
<span class="caption"></span>
<span class="attribution"><span class="license">Author provided</span></span>
</figcaption>
</figure>
<p>Based on current records, our big data driven prediction indicates that the final of the FIFA World Cup 2022 is set to be played between two-time winner France and their opponent, Argentina, on 18 December at <a href="https://www.qatar2022.qa/en/tournament/stadiums/lusail-stadium">Lusail Stadium</a>, Qatar. </p>
<p>France is predicted to be the first defending champions of the modern era since Brazil defended theirs in 1962, bringing joy to the country’s 67 million residents and its diaspora around the world. If these two teams make it to the final, then France is likely to be the favourite given their squad’s make-up and better defensive history in the course of the knockout stages.</p>
<p><strong>Final (predicted)</strong></p>
<figure class="align-center zoomable">
<a href="https://images.theconversation.com/files/500462/original/file-20221212-110747-xzjqgr.jpg?ixlib=rb-1.1.0&q=45&auto=format&w=1000&fit=clip"><img alt="" src="https://images.theconversation.com/files/500462/original/file-20221212-110747-xzjqgr.jpg?ixlib=rb-1.1.0&q=45&auto=format&w=754&fit=clip" srcset="https://images.theconversation.com/files/500462/original/file-20221212-110747-xzjqgr.jpg?ixlib=rb-1.1.0&q=45&auto=format&w=600&h=106&fit=crop&dpr=1 600w, https://images.theconversation.com/files/500462/original/file-20221212-110747-xzjqgr.jpg?ixlib=rb-1.1.0&q=30&auto=format&w=600&h=106&fit=crop&dpr=2 1200w, https://images.theconversation.com/files/500462/original/file-20221212-110747-xzjqgr.jpg?ixlib=rb-1.1.0&q=15&auto=format&w=600&h=106&fit=crop&dpr=3 1800w, https://images.theconversation.com/files/500462/original/file-20221212-110747-xzjqgr.jpg?ixlib=rb-1.1.0&q=45&auto=format&w=754&h=134&fit=crop&dpr=1 754w, https://images.theconversation.com/files/500462/original/file-20221212-110747-xzjqgr.jpg?ixlib=rb-1.1.0&q=30&auto=format&w=754&h=134&fit=crop&dpr=2 1508w, https://images.theconversation.com/files/500462/original/file-20221212-110747-xzjqgr.jpg?ixlib=rb-1.1.0&q=15&auto=format&w=754&h=134&fit=crop&dpr=3 2262w" sizes="(min-width: 1466px) 754px, (max-width: 599px) 100vw, (min-width: 600px) 600px, 237px"></a>
<figcaption>
<span class="caption"></span>
<span class="attribution"><a class="source" href="https://www.telegraph.co.uk/world-cup/2022-predictor-wallchart-football-bracket-predictions/">Developed using The Telegraph Wallchart</a>, <span class="license">Author provided</span></span>
</figcaption>
</figure>
<p>If the final games goes to the penalty stage, Argentina will hold the advantage given its recent records and its <a href="https://www.bbc.co.uk/sport/av/football/63925007">goalkeeper</a>. While a superb attacker, France’s Kylian Mbappé has only a <a href="https://theathletic.com/3978749/2022/12/09/world-cup-penalty-saves-success/">75% penalty conversion</a> success rate.</p>
<figure class="align-center zoomable">
<a href="https://images.theconversation.com/files/500464/original/file-20221212-108656-8w73ee.jpg?ixlib=rb-1.1.0&q=45&auto=format&w=1000&fit=clip"><img alt="" src="https://images.theconversation.com/files/500464/original/file-20221212-108656-8w73ee.jpg?ixlib=rb-1.1.0&q=45&auto=format&w=754&fit=clip" srcset="https://images.theconversation.com/files/500464/original/file-20221212-108656-8w73ee.jpg?ixlib=rb-1.1.0&q=45&auto=format&w=600&h=96&fit=crop&dpr=1 600w, https://images.theconversation.com/files/500464/original/file-20221212-108656-8w73ee.jpg?ixlib=rb-1.1.0&q=30&auto=format&w=600&h=96&fit=crop&dpr=2 1200w, https://images.theconversation.com/files/500464/original/file-20221212-108656-8w73ee.jpg?ixlib=rb-1.1.0&q=15&auto=format&w=600&h=96&fit=crop&dpr=3 1800w, https://images.theconversation.com/files/500464/original/file-20221212-108656-8w73ee.jpg?ixlib=rb-1.1.0&q=45&auto=format&w=754&h=120&fit=crop&dpr=1 754w, https://images.theconversation.com/files/500464/original/file-20221212-108656-8w73ee.jpg?ixlib=rb-1.1.0&q=30&auto=format&w=754&h=120&fit=crop&dpr=2 1508w, https://images.theconversation.com/files/500464/original/file-20221212-108656-8w73ee.jpg?ixlib=rb-1.1.0&q=15&auto=format&w=754&h=120&fit=crop&dpr=3 2262w" sizes="(min-width: 1466px) 754px, (max-width: 599px) 100vw, (min-width: 600px) 600px, 237px"></a>
<figcaption>
<span class="caption">Historic penalty shoot-out success from semi-final onwards.</span>
<span class="attribution"><span class="license">Author provided</span></span>
</figcaption>
</figure>
<p>Croatia is predicted to be third in the competition.</p>
<h2>Prediction validity and big surprises</h2>
<p>It is impossible to achieve 100% accuracy in predicting game outcomes, particularly in a tournament that is played at the highest level. Additional factors such as venue, host country weather, timing of the tournament, referee judgement, <a href="https://www.fifa.com/technical/football-technology/football-technologies-and-innovations-at-the-fifa-world-cup-2022/video-assistant-referee-var">video assistant referee</a> (VAR) interventions, squad formation, in-game tactical switches, and player concentration and stamina all play a huge role in producing the final outcome. These elements are relatively new to sports science and we are still unsure about how to apply them as influential statistical factors in an algorithm. </p>
<p>For example, VAR played a major role in Argentina’s <a href="https://www.bbc.com/sport/live/football/60976486">shock defeat against Saudi Arabia</a> and may eventually cause to Leonel Messi never lifting the world cup in his lifetime. Similarly, <a href="https://news.sky.com/story/japan-beat-germany-in-second-world-cup-shock-result-12753940">Japan’s victory over Germany</a> was a result of in-game tactics that the German players may not have expected. This World Cup promises to be exciting with lots of hidden surprises. We will have to wait until 18 December to find out who will be raising the trophy.</p><img src="https://counter.theconversation.com/content/196418/count.gif" alt="The Conversation" width="1" height="1" />
<p class="fine-print"><em><span>Les auteurs ne travaillent pas, ne conseillent pas, ne possèdent pas de parts, ne reçoivent pas de fonds d'une organisation qui pourrait tirer profit de cet article, et n'ont déclaré aucune autre affiliation que leur organisme de recherche.</span></em></p>After Brazil’s exit, who might be on course to win in 2022? Experts crunch 150 years of big data to predict the winner.Ronnie Das, Associate professor in Digital and Data Science, AudenciaWasim Ahmed, Senior Lecturer in Digital Business, University of StirlingLicensed as Creative Commons – attribution, no derivatives.tag:theconversation.com,2011:article/1926022022-11-24T19:05:43Z2022-11-24T19:05:43ZFriday essay: shaping history – why I spent ten years studying one Wikipedia article<figure><img src="https://images.theconversation.com/files/496330/original/file-20221121-9586-h0rj85.jpg?ixlib=rb-1.1.0&rect=6%2C134%2C4083%2C2323&q=45&auto=format&w=496&fit=clip" /><figcaption><span class="caption">An Egyptian woman takes part in a demonstration in Cairo, 25 January, 2011.</span> <span class="attribution"><span class="source">Amel Pain/EPA</span></span></figcaption></figure><p>In mid-July 2008, I arrived in hot and sticky Alexandria. I had travelled to Egypt to attend Wikimania. As the name suggests, Wikimania is an event for those who share an all-consuming passion for the wiki. But not just any wiki … the most important wiki of all: Wikipedia – the online encyclopedia. </p>
<p>This annual conference for Wikipedians (Wikipedia’s volunteer editors) is a chance to celebrate the project, discuss important issues, and geek out on wiki lore.</p>
<p>I was one of 650 attendees from 45 countries that year. But the conference (held in the Bibliotheca Alexandrina, an attempt to revive the Great Library of Alexandria) had been mired in controversy. There were calls to boycott the event because of Egypt’s censorship and imprisonment of bloggers. In his opening speech, Wikipedia co-founder Jimmy Wales highlighted the case of Abdel Kareem Nabil, a former university student sentenced to four years in prison on charges of insulting Islam and Egypt’s then President Hosni Mubarak, and inciting sectarian strife.</p>
<figure class="align-right zoomable">
<a href="https://images.theconversation.com/files/495572/original/file-20221116-16-5617k3.jpg?ixlib=rb-1.1.0&q=45&auto=format&w=1000&fit=clip"><img alt="" src="https://images.theconversation.com/files/495572/original/file-20221116-16-5617k3.jpg?ixlib=rb-1.1.0&q=45&auto=format&w=237&fit=clip" srcset="https://images.theconversation.com/files/495572/original/file-20221116-16-5617k3.jpg?ixlib=rb-1.1.0&q=45&auto=format&w=600&h=351&fit=crop&dpr=1 600w, https://images.theconversation.com/files/495572/original/file-20221116-16-5617k3.jpg?ixlib=rb-1.1.0&q=30&auto=format&w=600&h=351&fit=crop&dpr=2 1200w, https://images.theconversation.com/files/495572/original/file-20221116-16-5617k3.jpg?ixlib=rb-1.1.0&q=15&auto=format&w=600&h=351&fit=crop&dpr=3 1800w, https://images.theconversation.com/files/495572/original/file-20221116-16-5617k3.jpg?ixlib=rb-1.1.0&q=45&auto=format&w=754&h=441&fit=crop&dpr=1 754w, https://images.theconversation.com/files/495572/original/file-20221116-16-5617k3.jpg?ixlib=rb-1.1.0&q=30&auto=format&w=754&h=441&fit=crop&dpr=2 1508w, https://images.theconversation.com/files/495572/original/file-20221116-16-5617k3.jpg?ixlib=rb-1.1.0&q=15&auto=format&w=754&h=441&fit=crop&dpr=3 2262w" sizes="(min-width: 1466px) 754px, (max-width: 599px) 100vw, (min-width: 600px) 600px, 237px"></a>
<figcaption>
<span class="caption">Wikipedia co-founder Jimmy Wales pictured in 2008.</span>
<span class="attribution"><span class="source">Yonhap/AAP</span></span>
</figcaption>
</figure>
<p>Although some governments tried to impede free speech, Wales said, this was pointless in the age of the internet, where people could share ideas on platforms like Wikipedia.</p>
<p>“Kareem Amer has become a cause around the world,” he said, showing Nabil’s English Wikipedia page on the screen. “Not the best strategy for keeping his ideas out of the public eye.”</p>
<p>Two and a half years on, in late January 2011, Egyptians took to the streets to demand the end of authoritarian rule. Less than two weeks after protests erupted, Egypt’s autocrat president Mubarak resigned. Some were calling this “the Facebook revolution,” others a “Twitter revolution”. </p>
<p>Sadly it was to be short-lived. In 2013 Egyptian army chief General Abdel Fattah al-Sisi took over in a coup. He still rules today and <a href="https://www.theguardian.com/environment/2022/oct/18/greenwashing-police-state-egypt-cop27-masquerade-naomi-klein-climate-crisis">has imprisoned</a> an estimated 60,000 political prisoners, including those advocating democracy and free speech. </p>
<figure class="align-left zoomable">
<a href="https://images.theconversation.com/files/495563/original/file-20221116-11-mv7o0r.jpg?ixlib=rb-1.1.0&q=45&auto=format&w=1000&fit=clip"><img alt="" src="https://images.theconversation.com/files/495563/original/file-20221116-11-mv7o0r.jpg?ixlib=rb-1.1.0&q=45&auto=format&w=237&fit=clip" srcset="https://images.theconversation.com/files/495563/original/file-20221116-11-mv7o0r.jpg?ixlib=rb-1.1.0&q=45&auto=format&w=600&h=834&fit=crop&dpr=1 600w, https://images.theconversation.com/files/495563/original/file-20221116-11-mv7o0r.jpg?ixlib=rb-1.1.0&q=30&auto=format&w=600&h=834&fit=crop&dpr=2 1200w, https://images.theconversation.com/files/495563/original/file-20221116-11-mv7o0r.jpg?ixlib=rb-1.1.0&q=15&auto=format&w=600&h=834&fit=crop&dpr=3 1800w, https://images.theconversation.com/files/495563/original/file-20221116-11-mv7o0r.jpg?ixlib=rb-1.1.0&q=45&auto=format&w=754&h=1048&fit=crop&dpr=1 754w, https://images.theconversation.com/files/495563/original/file-20221116-11-mv7o0r.jpg?ixlib=rb-1.1.0&q=30&auto=format&w=754&h=1048&fit=crop&dpr=2 1508w, https://images.theconversation.com/files/495563/original/file-20221116-11-mv7o0r.jpg?ixlib=rb-1.1.0&q=15&auto=format&w=754&h=1048&fit=crop&dpr=3 2262w" sizes="(min-width: 1466px) 754px, (max-width: 599px) 100vw, (min-width: 600px) 600px, 237px"></a>
<figcaption>
<span class="caption">Police stand over a woman protesting on January 25, 2011.</span>
<span class="attribution"><span class="source">Amel Pain/EPA</span></span>
</figcaption>
</figure>
<p>But at the end of January 2011, as Mubarak still clung to power, the Israeli Wikipedian Dror Kamir wrote a startling message to a mailing list about Wikipedia’s role in the Egyptian protests.</p>
<p>Kamir pointed out that the first draft of the article about the Egyptian revolution on English Wikipedia had been published at 3:26pm local time, just hours after the first protests began. An Egyptian democracy activist and Wikipedian with the username The Egyptian Liberal had published this article, apparently to influence public opinion. “It almost seems as if the article preceded the actual events,” he wrote.</p>
<p>To Kamir, this demonstrated that Wikipedia “was losing its encyclopedic characteristics”. Wikipedians pride themselves on neutrality. Neutral point of view (or NPOV), is a core content policy. Editors are called to merely summarise reliable sources rather than offering their own original analysis. Policy determines that Wikipedians should follow public opinion rather than lead it. </p>
<p>In <a href="https://anduraru.wordpress.com/2011/06/25/parallel-online-and-real-world-egyptian-revolutions-or-wikipedias-tahrir-square/">a later blog post</a>, Kamir argued Wikipedia had clearly played a significant role in the events of January 2011, but “who is going to remember…?” </p>
<hr>
<p>
<em>
<strong>
Read more:
<a href="https://theconversation.com/cop27-shines-light-on-civil-liberties-in-egypt-but-itll-take-work-to-achieve-real-freedom-194407">COP27 shines light on civil liberties in Egypt, but it'll take work to achieve real freedom</a>
</strong>
</em>
</p>
<hr>
<h2>Remembering the history-makers</h2>
<p>In the coming months and years, I tried to do just that: documenting how Wikipedians wrote the story of the Egyptian revolution and whether, in doing so, they influenced the revolution itself.</p>
<p>It has been over a decade since I started studying this <a href="https://en.wikipedia.org/wiki/2011_Egyptian_revolution">single article on English Wikipedia about the 2011 revolution</a>. At the time of writing, it runs to almost 13,000 words and more than 400 citations.</p>
<p>Catalytic events have always been influenced by their mediation. But few had tried to understand Wikipedia’s role in history-making. When they did, they tended to present Wikipedia as hallowed ground where consensus is reached among a myriad alternative views.</p>
<p>The most important thing I have learned over this time is the truly subversive role of Wikipedia. Though the Egyptian revolution sputtered out, what I have gleaned from this example has a bearing on other history-making events playing out on Wikipedia now – from the war in Ukraine to the independence movement of Taiwan.</p>
<p>Wales was right when he gave that prescient speech. Wikipedia tends to be ignored because it is supposedly “neutral”. One of the world’s most popular platforms, maintained by <a href="https://wikimediafoundation.org/">a nonprofit organisation</a>, its mirage of neutrality is sustained by the idea that individuals may be biased but all crowds are wise.</p>
<p>Wikipedia supposedly reflects “common knowledge” and “collective memory”. But there are many different ways of seeing the world. There will always be an inevitable conflict between those tasked with its representation, especially when the risks and rewards are so great. How, then, did editors of the Egyptian article resolve these differences? What kind of history is the result?</p>
<figure class="align-center zoomable">
<a href="https://images.theconversation.com/files/495573/original/file-20221116-20-8me7a6.jpg?ixlib=rb-1.1.0&q=45&auto=format&w=1000&fit=clip"><img alt="A Wikipedia page." src="https://images.theconversation.com/files/495573/original/file-20221116-20-8me7a6.jpg?ixlib=rb-1.1.0&q=45&auto=format&w=754&fit=clip" srcset="https://images.theconversation.com/files/495573/original/file-20221116-20-8me7a6.jpg?ixlib=rb-1.1.0&q=45&auto=format&w=600&h=400&fit=crop&dpr=1 600w, https://images.theconversation.com/files/495573/original/file-20221116-20-8me7a6.jpg?ixlib=rb-1.1.0&q=30&auto=format&w=600&h=400&fit=crop&dpr=2 1200w, https://images.theconversation.com/files/495573/original/file-20221116-20-8me7a6.jpg?ixlib=rb-1.1.0&q=15&auto=format&w=600&h=400&fit=crop&dpr=3 1800w, https://images.theconversation.com/files/495573/original/file-20221116-20-8me7a6.jpg?ixlib=rb-1.1.0&q=45&auto=format&w=754&h=502&fit=crop&dpr=1 754w, https://images.theconversation.com/files/495573/original/file-20221116-20-8me7a6.jpg?ixlib=rb-1.1.0&q=30&auto=format&w=754&h=502&fit=crop&dpr=2 1508w, https://images.theconversation.com/files/495573/original/file-20221116-20-8me7a6.jpg?ixlib=rb-1.1.0&q=15&auto=format&w=754&h=502&fit=crop&dpr=3 2262w" sizes="(min-width: 1466px) 754px, (max-width: 599px) 100vw, (min-width: 600px) 600px, 237px"></a>
<figcaption>
<span class="caption">Wikipedia supposedly reflects ‘common knowledge’ and ‘collective memory’.</span>
<span class="attribution"><span class="source">Shutterstock</span></span>
</figcaption>
</figure>
<p>A typical Wikipedia article is put together by Wikipedians – the volunteer editors who are committed to Wikipedia’s long-term maintenance. Anyone can be a Wikipedian, as long as you abide by <a href="https://en.wikipedia.org/wiki/Wikipedia:Policies_and_guidelines">the rules of the project</a> (many have a long history with the site). Wikipedians tend to use pseudonyms rather than their real names – there has been no policy requiring them to identify themselves.</p>
<p>As well as Wikipedians, entries are generally open to anyone else to edit. Many Wikipedians volunteer to watch over articles, receiving an alert when changes have been made to assess them.</p>
<p>A <a href="https://en.wikipedia.org/w/index.php?title=Wikipedia:What_Wikipedia_is_not#Wikipedia_is_not_a_crystal_ball">key Wikipedia rule</a> is that Wikipedia is not “a crystal ball”. The rule stipulates that Wikipedians should not write about events until their significance is generally known or before the event has concluded.</p>
<p>Soon after the revolution in 2011, I began analysing countless “talk page” discussions where Wikipedia editors discussed the reliability of sources, how to source free images and how to best summarise these events. (These discussions take place in a tab next to the article labelled “talk”.) </p>
<p>Over the next decade, I reviewed hundreds of edits, and interviewed leading editors. These included The Egyptian Liberal, a university student in his twenties, and Ocaasi, a US-based college graduate in his late twenties. Ocaasi, who suffered from anxiety and agoraphobia, told me he was editing Wikipedia obsessively at the time of the protests <a href="https://wikipedia20.mitpress.mit.edu/pub/k3dp712w/release/17">while sitting in his bathtub in a Philadelphia attic</a>. </p>
<figure class="align-right zoomable">
<a href="https://images.theconversation.com/files/495780/original/file-20221117-27-58ojhi.png?ixlib=rb-1.1.0&q=45&auto=format&w=1000&fit=clip"><img alt="Head shot of a man." src="https://images.theconversation.com/files/495780/original/file-20221117-27-58ojhi.png?ixlib=rb-1.1.0&q=45&auto=format&w=237&fit=clip" srcset="https://images.theconversation.com/files/495780/original/file-20221117-27-58ojhi.png?ixlib=rb-1.1.0&q=45&auto=format&w=600&h=611&fit=crop&dpr=1 600w, https://images.theconversation.com/files/495780/original/file-20221117-27-58ojhi.png?ixlib=rb-1.1.0&q=30&auto=format&w=600&h=611&fit=crop&dpr=2 1200w, https://images.theconversation.com/files/495780/original/file-20221117-27-58ojhi.png?ixlib=rb-1.1.0&q=15&auto=format&w=600&h=611&fit=crop&dpr=3 1800w, https://images.theconversation.com/files/495780/original/file-20221117-27-58ojhi.png?ixlib=rb-1.1.0&q=45&auto=format&w=754&h=768&fit=crop&dpr=1 754w, https://images.theconversation.com/files/495780/original/file-20221117-27-58ojhi.png?ixlib=rb-1.1.0&q=30&auto=format&w=754&h=768&fit=crop&dpr=2 1508w, https://images.theconversation.com/files/495780/original/file-20221117-27-58ojhi.png?ixlib=rb-1.1.0&q=15&auto=format&w=754&h=768&fit=crop&dpr=3 2262w" sizes="(min-width: 1466px) 754px, (max-width: 599px) 100vw, (min-width: 600px) 600px, 237px"></a>
<figcaption>
<span class="caption">Ocaasi (Jake Orlowitz).</span>
<span class="attribution"><span class="source">Author provided</span></span>
</figcaption>
</figure>
<p>Rather than rational negotiation and broad consensus, I learned that Wikipedia articles about historic events are often the result of passionate struggle over representing what happened to whom and its consequences.</p>
<p>I learned about the importance of Wikipedians themselves in shaping the narrative into which individual facts were made to fit. Wikipedians shaped the representation of the event not by inserting falsities but rather by framing and selecting facts that supported certain narratives rather than others. </p>
<p>The Wikipedians moved quickly to create a new article on English Wikipedia when crowds first swarmed into Tahrir Square on January 25, 2011, defending it from possible attack from sceptics arguing it was too soon to be covering events. They bolstered the article’s authority by quickly adding citations to source the evidence for the unfolding protests. </p>
<p>This first move was successful in determining that the protests were important enough to warrant their own article early on. </p>
<hr>
<p>
<em>
<strong>
Read more:
<a href="https://theconversation.com/mubarak-a-man-who-built-on-his-talent-for-self-promotion-while-stifling-opposition-132565">Mubarak: a man who built on his talent for self-promotion while stifling opposition</a>
</strong>
</em>
</p>
<hr>
<h2>Writing the revolution into being</h2>
<p>After the first 24 hours in the life of the article, it had been edited 130 times. Forty two editors had joined The Egyptian Liberal including two longtime Wikipedians, Dragons flight, a physicist educated at UC Berkeley now living in Switzerland and Heroeswithmetaphors, who has made over 18,000 contributions to articles on multiple topics.</p>
<p>Editors settled into a routine – with American editors handing over to those in Egypt and elsewhere when they went to sleep. For Ocaasi, it was a galvanizing moment. “Everything before that on Wikipedia was just playing around and this was not,” he told me. </p>
<blockquote>
<p>It was also when my innocence about Wikipedia ended. It wasn’t just a hobby or escape […] There were hundreds of thousands of people reading the article and I knew that. There was a profound sense of responsibility […] I thought the world mattered so much those days and I thought I could play a part – not in an activist sense but by documenting what was happening. </p>
</blockquote>
<p>As the violent protests continued, experienced editors resisted attempts by newcomers to continuously change the article’s title from “protests” to “revolution”. A move of this significance requires consensus from editors on the talk page.</p>
<figure class="align-center zoomable">
<a href="https://images.theconversation.com/files/495570/original/file-20221116-22-9nhhbn.jpg?ixlib=rb-1.1.0&q=45&auto=format&w=1000&fit=clip"><img alt="A demonstration in Cairo." src="https://images.theconversation.com/files/495570/original/file-20221116-22-9nhhbn.jpg?ixlib=rb-1.1.0&q=45&auto=format&w=754&fit=clip" srcset="https://images.theconversation.com/files/495570/original/file-20221116-22-9nhhbn.jpg?ixlib=rb-1.1.0&q=45&auto=format&w=600&h=382&fit=crop&dpr=1 600w, https://images.theconversation.com/files/495570/original/file-20221116-22-9nhhbn.jpg?ixlib=rb-1.1.0&q=30&auto=format&w=600&h=382&fit=crop&dpr=2 1200w, https://images.theconversation.com/files/495570/original/file-20221116-22-9nhhbn.jpg?ixlib=rb-1.1.0&q=15&auto=format&w=600&h=382&fit=crop&dpr=3 1800w, https://images.theconversation.com/files/495570/original/file-20221116-22-9nhhbn.jpg?ixlib=rb-1.1.0&q=45&auto=format&w=754&h=480&fit=crop&dpr=1 754w, https://images.theconversation.com/files/495570/original/file-20221116-22-9nhhbn.jpg?ixlib=rb-1.1.0&q=30&auto=format&w=754&h=480&fit=crop&dpr=2 1508w, https://images.theconversation.com/files/495570/original/file-20221116-22-9nhhbn.jpg?ixlib=rb-1.1.0&q=15&auto=format&w=754&h=480&fit=crop&dpr=3 2262w" sizes="(min-width: 1466px) 754px, (max-width: 599px) 100vw, (min-width: 600px) 600px, 237px"></a>
<figcaption>
<span class="caption">A tank stands amid crowds as protesters gather on Tahrir Square in Cairo, Egypt, 1 February 2011.</span>
<span class="attribution"><span class="source">Hannibal Hanschke/EPA</span></span>
</figcaption>
</figure>
<p>But within minutes of Mubarak’s resignation at 4pm on February 11, a large crowd of Wikipedia editors again tried to change key facts to reclassify the article to “revolution.” In the hour after Mubarak resigned, the number of readers accessing the page tripled from about 4,000 to 12,500. It was being edited every two minutes in the following hours, as three experienced Wikipedians struggled to hold back the flood of editors attempting to make significant changes before consensus had been reached about the title. </p>
<p>While this was happening, a discussion began on the talk page, with editors asked to weigh in on whether the title of the article should be changed. But an editor, Tariqabjotu, made the change just two hours after Mubarak’s resignation – long before the discussion had run its course. </p>
<p>At this time also, the article on the Tunisian protests, which had unseated long-time President Zine El Abidine Ben Ali the month before, was still merely named as an uprising. Six hours after Mubarek’s resignation, another editor, Knowledgekid87, moved from editing the Egyptian article to the Tunisian one, changing its name to “Tunisian revolution”. This reinforced the Egyptian title change.</p>
<p>On February 11 alone, 125,000 readers accessed the Egyptian article. The number of editors working on it more than tripled from 25 to 84. Many were new editors from the United States, UK, Canada, the Netherlands, Portugal and Singapore, overwhelming those who had been editing it consistently from the beginning.</p>
<p>Other editors in the crowd repeatedly changed the date of the events in the infobox (the small fact box on the right hand side of a Wikipedia article) from “25 January – ongoing” to “25 January – 11 February”. They did this to cement the idea that the protests were over and revolution had been achieved. </p>
<p>In my interviews with Ocaasi, he reflected on how editors surrendered to the momentousness of the occasion. Any effort to resist changes to the article’s title would have been swimming against a tide of editors, one of whom declared that Wikipedia shouldn’t “deny history.” </p>
<figure class="align-center zoomable">
<a href="https://images.theconversation.com/files/495569/original/file-20221116-20-yjn9z8.jpg?ixlib=rb-1.1.0&rect=0%2C17%2C4000%2C2580&q=45&auto=format&w=1000&fit=clip"><img alt="A crowd in Cairo" src="https://images.theconversation.com/files/495569/original/file-20221116-20-yjn9z8.jpg?ixlib=rb-1.1.0&rect=0%2C17%2C4000%2C2580&q=45&auto=format&w=754&fit=clip" srcset="https://images.theconversation.com/files/495569/original/file-20221116-20-yjn9z8.jpg?ixlib=rb-1.1.0&q=45&auto=format&w=600&h=391&fit=crop&dpr=1 600w, https://images.theconversation.com/files/495569/original/file-20221116-20-yjn9z8.jpg?ixlib=rb-1.1.0&q=30&auto=format&w=600&h=391&fit=crop&dpr=2 1200w, https://images.theconversation.com/files/495569/original/file-20221116-20-yjn9z8.jpg?ixlib=rb-1.1.0&q=15&auto=format&w=600&h=391&fit=crop&dpr=3 1800w, https://images.theconversation.com/files/495569/original/file-20221116-20-yjn9z8.jpg?ixlib=rb-1.1.0&q=45&auto=format&w=754&h=491&fit=crop&dpr=1 754w, https://images.theconversation.com/files/495569/original/file-20221116-20-yjn9z8.jpg?ixlib=rb-1.1.0&q=30&auto=format&w=754&h=491&fit=crop&dpr=2 1508w, https://images.theconversation.com/files/495569/original/file-20221116-20-yjn9z8.jpg?ixlib=rb-1.1.0&q=15&auto=format&w=754&h=491&fit=crop&dpr=3 2262w" sizes="(min-width: 1466px) 754px, (max-width: 599px) 100vw, (min-width: 600px) 600px, 237px"></a>
<figcaption>
<span class="caption">Tens of thousands of Egyptians pray and celebrate the fall of the regime of former President Hosni Mubarak on February 18 2011.</span>
<span class="attribution"><span class="source">Ben Curtis/AP</span></span>
</figcaption>
</figure>
<p>The crowd centred their activity on the infobox and the page name. These elements are the most important parts of a Wikipedia article because they present summarised facts that appear authoritative and stable. These facts have always been prioritised by Google and other search engines’ algorithms, which often place Wikipedia at the top of search results. But the infobox came to matter even more the year after the Egyptian revolution. </p>
<p>In 2012, Google <a href="https://www.blog.google/products/search/introducing-knowledge-graph-things-not/">announced</a> a major new project that would build a massive database of facts built from “public” information sources such as Wikipedia and the CIA World Factbook. </p>
<p>Google’s algorithms selectively extract facts from Wikipedia’s infoboxes, divorcing them from the context in which they originated. Sources and citations are often removed. The facts appear more stable than they are on Wikipedia, where they are flanked by breaking news warnings and “citation needed” tags. Wikipedians have no control over Google’s process. </p>
<p>Over a decade after the 2011 Egyptian revolution, Wikipedia is still the authority for facts about the event. If you ask Google, Bing or Yahoo what happened in Egypt in 2011, they will present facts extracted from the English Wikipedia article. </p>
<p>But Google and other platforms extract them automatically and without understanding or debate. The result is a representation of capitalist logic embedded in the machines that have been programmed not to serve public meaning-making but rather to feed revenue sources. </p>
<p>For the past few years, Google’s knowledge panel about the revolution has contained the words, “Deaths section below” after facts about numbers killed during the revolution. This is material lifted from Wikipedia but not linked to further information – so it becomes a meaningless phrase. It shows how Wikipedians have lost control over some of the information they carefully provide. Yet many more people will view this material now in a search engine rather than on Wikipedia. </p>
<hr>
<p>
<em>
<strong>
Read more:
<a href="https://theconversation.com/egypt-hopes-for-democratic-future-die-as-al-sisi-marches-country-towards-dictatorship-with-parliaments-blessing-113491">Egypt: hopes for democratic future die as al-Sisi marches country towards dictatorship – with parliament's blessing</a>
</strong>
</em>
</p>
<hr>
<h2>A struggle for power</h2>
<p>Popular accounts like <a href="https://en.wikipedia.org/wiki/The_Wisdom_of_Crowds">The Wisdom of Crowds</a> and <a href="https://www.wired.com/2008/06/pb-theory/">the End of Theory</a> present both crowds and algorithms as sources of truth and neutrality. By such accounts, crowds supposedly smooth out one anothers’ biases or ignorance and Big Data enables accuracy because of our access to huge datasets. </p>
<figure class="align-left zoomable">
<a href="https://images.theconversation.com/files/495574/original/file-20221116-16-rnqjwn.jpg?ixlib=rb-1.1.0&q=45&auto=format&w=1000&fit=clip"><img alt="" src="https://images.theconversation.com/files/495574/original/file-20221116-16-rnqjwn.jpg?ixlib=rb-1.1.0&q=45&auto=format&w=237&fit=clip" srcset="https://images.theconversation.com/files/495574/original/file-20221116-16-rnqjwn.jpg?ixlib=rb-1.1.0&q=45&auto=format&w=600&h=899&fit=crop&dpr=1 600w, https://images.theconversation.com/files/495574/original/file-20221116-16-rnqjwn.jpg?ixlib=rb-1.1.0&q=30&auto=format&w=600&h=899&fit=crop&dpr=2 1200w, https://images.theconversation.com/files/495574/original/file-20221116-16-rnqjwn.jpg?ixlib=rb-1.1.0&q=15&auto=format&w=600&h=899&fit=crop&dpr=3 1800w, https://images.theconversation.com/files/495574/original/file-20221116-16-rnqjwn.jpg?ixlib=rb-1.1.0&q=45&auto=format&w=754&h=1130&fit=crop&dpr=1 754w, https://images.theconversation.com/files/495574/original/file-20221116-16-rnqjwn.jpg?ixlib=rb-1.1.0&q=30&auto=format&w=754&h=1130&fit=crop&dpr=2 1508w, https://images.theconversation.com/files/495574/original/file-20221116-16-rnqjwn.jpg?ixlib=rb-1.1.0&q=15&auto=format&w=754&h=1130&fit=crop&dpr=3 2262w" sizes="(min-width: 1466px) 754px, (max-width: 599px) 100vw, (min-width: 600px) 600px, 237px"></a>
<figcaption>
<span class="caption"></span>
</figcaption>
</figure>
<p>But I discovered a passion and feverish anticipation of revolution in Egypt from the very first entry on it, just hours after the protests began on January 25. Rather than rational consensus among dispassionate observers, Wikipedia mirrored the passion, emotion and violence of Tahrir Square. </p>
<p>Did Wikipedia shape the political events at the time, as suggested by Kamir? Ultimately, the story of this Wikipedia entry reiterates how young people (the leading Wikipedia editors) were able to win the information war in Egypt but not transform the government. Most of the article’s editors were people in favour of the revolution.</p>
<p>Nevertheless, Wikipedia articles about political events are important battlegrounds for interest groups vying for control over the historical record. Their impact lives on, courtesy of search engines’ algorithms and the global reach of the site itself. And such struggles for power are no doubt happening, elsewhere, in other Wikipedia articles today.</p>
<p><em>Writing the Revolution: Wikipedia and the Survival of Facts in the Digital Age is published by <a href="https://mitpress.mit.edu/9780262046299/writing-the-revolution/">MIT Press</a>.</em></p><img src="https://counter.theconversation.com/content/192602/count.gif" alt="The Conversation" width="1" height="1" />
<p class="fine-print"><em><span>Heather Ford receives funding from the Australian Research Council.</span></em></p>How are Wikipedia pages about contentious events put together? Heather Ford discovered a hotbed of passion, a rotating pack of editors and a struggle for power behind its mirage of neutrality.Heather Ford, Associate Professor, University of Technology SydneyLicensed as Creative Commons – attribution, no derivatives.tag:theconversation.com,2011:article/1946012022-11-18T16:05:50Z2022-11-18T16:05:50ZWhat the world would lose with the demise of Twitter: Valuable eyewitness accounts and raw data on human behavior, as well as a habitat for trolls<figure><img src="https://images.theconversation.com/files/495986/original/file-20221117-15-s4q1at.jpg?ixlib=rb-1.1.0&rect=0%2C0%2C6000%2C3997&q=45&auto=format&w=496&fit=clip" /><figcaption><span class="caption">Twitter itself produces a lot of data that's available nowhere else.</span> <span class="attribution"><a class="source" href="https://www.gettyimages.com/detail/news-photo/the-twitter-logo-is-seen-in-this-photo-illustration-in-news-photo/1244760225">STR/NurPhoto via Getty Images</a></span></figcaption></figure><p>What do a cybersecurity researcher building a system to generate alerts for detecting <a href="https://ieeexplore.ieee.org/abstract/document/7752338">security threats and vulnerabilities</a>, a wildfire watcher who <a href="https://www.wired.com/story/california-fire-twitter/">tracks the spread of forest fires</a>, and public health professionals <a href="https://www.ncbi.nlm.nih.gov/pmc/articles/PMC5308155/">trying to predict enrollment</a> in health insurance exchanges have in common? </p>
<p>They all rely on analyzing data from Twitter. </p>
<p>Twitter is a microblogging service, meaning it’s designed for sharing posts of short segments of text and embedded audio and video clips. The ease with which people can share information among millions of others worldwide on Twitter has made it very popular for real-time conversations. Whether it is people tweeting about their favorite sports teams, or organizations and public figures using Twitter to reach a mass audience, Twitter has been part of the collective record for over a decade. </p>
<p>The Twitter <a href="https://blog.twitter.com/en_us/a/2015/full-archive-search-api">archives allow for instant and complete access</a> to every public tweet, which has positioned Twitter both as a archive of collective human behavior and as a credentialing and fact-checking service on a global scale. As a <a href="https://scholar.google.com/citations?user=JpFHYKcAAAAJ&hl=en">researcher who studies social media</a>, I believe that these functions are very valuable for academics, policymakers and anyone using aggregate data to obtain insights into human behavior. </p>
<p>The proliferation of <a href="https://www.wired.com/story/twitter-blue-check-verification-buy-scams/">scams and brand impersonators</a>, the <a href="https://www.washingtonpost.com/technology/2022/11/14/twitter-fake-eli-lilly/">hemorrhaging of advertisers</a>, and <a href="https://www.nytimes.com/2022/11/17/technology/twitter-elon-musk-ftc.html">disarray within the company</a> call the future of the platform <a href="https://mashable.com/article/elon-musk-twitter-future">into question</a>. If Twitter were to go under, the loss would reverberate around the world. </p>
<h2>Analyzing human behavior</h2>
<p>With its massive trove of tweets, Twitter has provided new ways to quantify public discourse and new tools to map aggregate perceptions, and offers a window into large-scale human behavior. Such <a href="https://library.oapen.org/viewer/web/viewer.html?file=/bitstream/handle/20.500.12657/51412/9781003024583_10.4324_9781003024583-8.pdf">digital traces</a> or records of human activity allow researchers in fields ranging from social sciences to healthcare to analyze a variety of phenomena. </p>
<p>From open source intelligence to citizen science, Twitter has not only been a digital public square, but has also allowed researchers to infer attitudes that are difficult to detect through methods from traditional field research. For example, people’s willingness to pay for policies and services that address climate change has traditionally been measured through surveys of subjective well-being. Twitter sentiment data gives researchers and policymakers <a href="https://doi.org/10.1016/j.jpubeco.2020.104161">another tool for assessing these attitudes</a> in order to take more meaningful action on climate change. </p>
<p>Researchers in public health have found an association between tweeting about <a href="https://doi.org/10.2196%2F17196">HIV and incidence of HIV</a>, and have been able to <a href="https://doi.org/10.1371/journal.pone.0219550">measure sentiment at the neighborhood level</a> to assess the overall health of the people in those neighborhoods. </p>
<h2>Place and time</h2>
<p>Geotagged data from Twitter helps in a variety of fields such as <a href="https://doi.org/10.1371/journal.pone.0131469">urban land use</a> and <a href="https://doi.org/10.1080/24694452.2017.1421897">disaster resilience</a>. Being able to identify the locations for a set of tweets allows researchers to correlate information in the tweets with times and places – for example, correlating tweets and ZIP codes to <a href="https://theconversation.com/matching-tweets-to-zip-codes-can-spotlight-hot-spots-of-covid-19-vaccine-hesitancy-169596">identify hot spots of vaccine hesitancy</a>.</p>
<p>Twitter has been invaluable in the field of <a href="https://doi.org/10.1109/ICCCN49398.2020.9209602">open source intelligence</a> (OSINT), particularly for tracking down war crimes. OSINT uses crowdsourcing to identify the locations of photos and videos. In Ukraine, <a href="https://www.grid.news/story/global/2022/04/11/in-ukraine-war-crimes-are-being-captured-on-social-media/">human rights investigators have focused on using Twitter and TikTok</a> to search for evidence of abuses. </p>
<p>Open source intelligence has also been helpful for cutting through the fog of war. For example, OSINT analysts were quick to provide evidence that the missile that exploded in Przewodow, Poland near the Ukrainian border on Nov. 15, 2022 was likely an S-300 antiaircraft missile and unlikely a ballistic or cruise missile fired by Russia.</p>
<p><div data-react-class="Tweet" data-react-props="{"tweetId":"1592629251161075712"}"></div></p>
<h2>Credentialing and verification</h2>
<p>Although misinformation has been <a href="https://doi.org/10.1126/science.aap9559">disseminated far and wide on Twitter</a>, the platform also serves a role as a global verification mechanism. First, vast numbers of people use Twitter and other social media platforms. With crowdsourcing writ large, social media assumes the role of an authoritative information provider, reducing some of the uncertainty people face in searching for new information. The platforms perform a credentialing role that some scholars refer to as “<a href="https://www.google.com/books/edition/Media_Technologies/zeK2AgAAQBAJ">public relevance algorithms</a>,” in that they have replaced dedicated business or technical expertise in identifying what people need to know. </p>
<p>Another way has been official credentialing. Prior to Elon Musk’s takeover, Twitter’s verification method provided public figures with a blue check mark on their profiles, which served as a shortcut in establishing whether a source of a tweet <a href="https://www.politico.com/news/2022/11/13/washington-gets-increasingly-freaked-out-by-twitter-00066647">was who the person claimed to be</a>. </p>
<p>While problems such as <a href="https://doi.org/10.1126/science.aau2706">fake news</a>, <a href="https://doi.org/10.2105/AJPH.2020.305940">misinformation</a> and <a href="https://aclanthology.org/W18-5110/">hate speech</a> exist, the credentialing ability coupled with the vast number of people who use the platform in real time made Twitter a provider of credible information and a <a href="https://doi.org/10.1177/1940161214540942">fact-checker</a>. </p>
<h2>The digital public square</h2>
<p>Twitter’s dual role in fostering real-time communication and acting as an arbitrator of authoritative information is of crucial interest to academics, journalists and government agencies. During the pandemic, for example, many public health <a href="https://doi.org/10.2196/24883">agencies turned to Twitter</a> to promote behavior that mitigates the risk of infection. </p>
<p><div data-react-class="Tweet" data-react-props="{"tweetId":"1592676469305905152"}"></div></p>
<p>During disasters and emergencies, Twitter has been a great venue for <a href="https://doi.org/10.1016/j.ipm.2019.102107">crowdsourced eyewitness data</a>. During Hurricane Harvey, for example, researchers found that that users responded and interacted the <a href="https://doi.org/10.1007/s11069-020-04016-6">most with tweets from verified Twitter accounts</a>, and especially from government organizations. Official Twitter accounts helped in the rapid dissemination of information during <a href="https://doi.org/10.1016/j.chb.2015.06.044">a water contamination crisis</a> in West Virginia. Twitter data has also helped in <a href="https://doi.org/10.1016/j.trc.2021.102976">hurricane evacuations</a>. </p>
<p>Twitter has also been an important way for people with disabilities to participate in public discourse.</p>
<p><div data-react-class="Tweet" data-react-props="{"tweetId":"1593334136969666560"}"></div></p>
<p>Twitter’s real value has been in enabling people to connect with each other in real time and as an archive of collective behavior. Recognizing this, <a href="https://qz.com/1143475/the-un-is-the-international-organization-with-the-most-followers-on-twitter">international organizations</a>, <a href="https://fcw.com/workforce/2012/09/the-50-most-followed-agencies-on-twitter/206197/">government agencies</a> and <a href="https://icma.org/articles/article/top-local-government-twitter-users">local governments</a> have invested significant resources in using Twitter and have come to rely on the platform. Sen. Edward Markey has described Twitter as “<a href="https://www.politico.com/amp/news/2022/11/17/ed-markey-deep-dive-00069221">essential” to American society</a>. If Twitter were to collapse, there’s no clear replacement in sight.</p><img src="https://counter.theconversation.com/content/194601/count.gif" alt="The Conversation" width="1" height="1" />
<p class="fine-print"><em><span>Anjana Susarla does not work for, consult, own shares in or receive funding from any company or organization that would benefit from this article, and has disclosed no relevant affiliations beyond their academic appointment.</span></em></p>If Twitter were to go dark, with it would go a valuable source of data as well as a means of sharing information relied on by activists, journalists, public health officials and scientists.Anjana Susarla, Professor of Information Systems, Michigan State UniversityLicensed as Creative Commons – attribution, no derivatives.tag:theconversation.com,2011:article/1912662022-10-24T13:17:41Z2022-10-24T13:17:41ZArtificial intelligence is used for predictive policing in the US and UK – South Africa should embrace it, too<figure><img src="https://images.theconversation.com/files/486478/original/file-20220926-14-5pa015.jpg?ixlib=rb-1.1.0&q=45&auto=format&w=496&fit=clip" /><figcaption><span class="caption">Predictive policing may be a useful addition to traditional policing in contexts like South Africa.</span> <span class="attribution"><span class="source">Fani Mahuntsi/Gallo Images via Getty Images</span></span></figcaption></figure><p>In the 2002 movie Minority Report (based on a <a href="https://www.goodreads.com/book/show/581125.The_Minority_Report">short story</a> by Philip K Dick), director Steven Spielberg imagined a future in which three psychics can “see” murders before they happen. Their clairvoyance allows Tom Cruise and his “Precrime” police force to avert nearly all potential homicides.</p>
<p>Twenty years on, in the real world, scientists and law enforcement agencies are using data mining and machine learning to mimic those psychics. Such “<a href="https://www.science.org/content/article/can-predictive-policing-prevent-crime-it-happens">predictive policing</a>”, as it is called, is based on the fact that many crimes – and criminals – have <a href="https://www.taylorfrancis.com/chapters/edit/10.4324/9780203118214-13/crime-pattern-theory-paul-brantingham-patricia-brantingham">detectable patterns</a>.</p>
<p>Predictive policing has enjoyed some successes. In a <a href="https://www.tandfonline.com/doi/full/10.1080/01900692.2019.1575664">case study</a> in the US, one police department was able to reduce gun incidents by 47% over the typically gun-happy New Year’s Eve. <a href="https://www.ironsidegroup.com/case-study/predictive-policing-success-manchester-police-department/">Manchester police</a> in the UK were similarly able to predict and reduce robberies, burglaries and thefts from motor vehicles by double digits in the first 10 weeks of rolling out predictive measures.</p>
<p>Predictive policing has improved in leaps and bounds. <a href="https://link.springer.com/content/pdf/10.1007/978-3-642-40994-3_33.pdf">In the past</a>, humans had to manually pore over crime reports or filter through national crime databases. Now, in the age of <a href="https://www.igi-global.com/book/data-mining-trends-applications-criminal/146986">big data, data mining</a> and powerful computers, that process can be automated. </p>
<p>But merely finding information is not enough to deter crime. The data needs to be analysed to detect underlying patterns and relationships. Scientists deploy algorithms and mathematical models such as machine learning, which imitates the way humans learn, to extract useful information and insights from existing data. </p>
<p>Recently, we <a href="https://dl.acm.org/doi/fullHtml/10.1145/3488933.3488973">turned to</a> a mathematical method conceived in the 18th century to refine our approach. By tweaking an existing algorithm based on this method, we significantly improved its crime prediction rates.</p>
<p>This finding holds promise for applying predictive policing in under-resourced contexts like South Africa. This could help reduce crime levels – some of the highest in the world and <a href="https://www.dailymaverick.co.za/article/2022-06-03-crime-crisis-continues-in-first-quarter-of-2022-with-women-and-children-worst-affected">rising</a>. It’s a situation the country’s police force seems <a href="https://africacheck.org/infofinder/explore-facts/how-many-people-does-one-police-officer-serve-south-africa">ill-equipped</a> to curb.</p>
<h2>Marrying two different approaches</h2>
<p>Thomas Bayes was a British mathematician. His famed <a href="https://www.mathsisfun.com/data/bayes-theorem.html">Bayes’ theorem</a> essentially describes the probability of an event occurring based on some prior knowledge of conditions that may be related to that event. Today, Bayesian analysis is commonplace in fields as diverse as artificial intelligence, astrophysics, finance, gambling and weather forecasting. We fine-tuned the Naïve Bayes algorithm and <a href="https://dl.acm.org/doi/fullHtml/10.1145/3488933.3488973">put it to the test</a> as a crime predictor. </p>
<p>Bayesian analysis can use probability statements to answer research questions about unknown parameters of statistical models. For example, what is the probability that a suspect accused of a crime is guilty? But going deeper – like calculating how poker cards may unfold, or how humans (especially humans with criminal intent) will act – requires increasingly sophisticated technologies and algorithms. </p>
<p><a href="https://dl.acm.org/doi/fullHtml/10.1145/3488933.3488973">Our research</a> built on the Naïve Bayes algorithm or classifier, a popular supervised machine learning algorithm, for <a href="https://dl.acm.org/doi/fullHtml/10.1145/3488933.3488973">crime prediction</a>. </p>
<p>Naïve Bayes starts on the premise that features – the variables that serve as input – are conditionally independent, meaning that the presence of one feature does not affect the others.</p>
<p>We fine-tuned the Naïve Bayes algorithm by marrying it with another algorithm known as <a href="https://machinelearningmastery.com/rfe-feature-selection-in-python/">Recursive Feature Elimination</a>. This tool assists in selecting the more significant features in a dataset and removing the weaker ones, with the objective of improving the results.</p>
<p>We then applied our finessed algorithm to a popular experimental dataset extracted from the Chicago Police Department’s <a href="https://home.chicagopolice.org/services/clearmap-application/">CLEAR</a> (Citizen Law Enforcement Analysis and Reporting) system, which has been used to predict and reduce crime in that American city. That dataset has been applied globally because of the rich data it contains: it provides incident-level crime data, registered offenders, community concerns, and locations of police stations in the city.</p>
<p>We compared the results of our enhanced Naïve Bayes against that of the original Naïve Bayes, as well as against other predictive algorithms such as Random Forests and Extremely Randomized Trees (algorithms we have <a href="https://dl.acm.org/doi/fullHtml/10.1145/3488933.3488972">also worked</a> on for crime prediction). We found that we could improve on the predictions of the Naïve Bayes by about 30%, and could either match or improve on the predictions of the other algorithms.</p>
<h2>Data and bias</h2>
<p>While our model holds promise, there’s one element that’s sorely lacking in applying it to South African contexts: data. As the Chicago CLEAR system illustrates, predictive models work best when you have lots of relevant data to work with. But South Africa’s police force has historically been very tight-fisted with its data, perhaps due to confidentiality issues. I ran into this problem in my <a href="https://open.uct.ac.za/handle/11427/25319">doctoral research</a> on detecting and mapping crime series.</p>
<p>This is slowly shifting. We are currently running a small case study in Bellville, a suburb about 20km from Cape Town’s central business district and the area in which our university is located, using the <a href="https://www.kaggle.com/datasets/slwessels/crime-statistics-for-south-africa">South African Police Service data</a> for predictive policing.</p>
<p>None of this is to suggest that predictive policing alone will solve South Africa’s crime problem. Predictive algorithms and policing are not without their flaws. Even the psychics in Minority Report, it turned out, were not error-free. Fears that these algorithms may simply reinforce racial biases, for instance, have been raised both in <a href="https://www.dailymaverick.co.za/article/2022-08-02-sa-police-may-be-jumping-the-gun-by-implementing-new-crimefighting-technologies/">South Africa</a> and <a href="https://www.science.org/content/article/can-predictive-policing-prevent-crime-it-happens">elsewhere</a>.</p>
<p>But we believe that, with continuous technological improvement, predictive policing could play an important role in bolstering the police’s responsiveness and may be a small step towards improving public confidence in the police.</p>
<p><em>Dr Olasupo Ajayi of the Department of Computer Science at the University of the Western Cape and Mr Sphamandla May, a master’s student in the department, co-authored this article and the research it’s based on.</em></p><img src="https://counter.theconversation.com/content/191266/count.gif" alt="The Conversation" width="1" height="1" />
<p class="fine-print"><em><span>Omowunmi Isafiade does not work for, consult, own shares in or receive funding from any company or organisation that would benefit from this article, and has disclosed no relevant affiliations beyond their academic appointment.</span></em></p>Predictive policing has improved in leaps and bounds and become increasingly automated thanks to big data, data mining and powerful computers.Omowunmi Isafiade, Senior Lecturer in Computer Science, University of the Western CapeLicensed as Creative Commons – attribution, no derivatives.tag:theconversation.com,2011:article/1890522022-10-10T11:59:41Z2022-10-10T11:59:41ZCan big data really predict what makes a song popular?<figure><img src="https://images.theconversation.com/files/487407/original/file-20220929-14-ontc5y.jpg?ixlib=rb-1.1.0&rect=0%2C574%2C7988%2C4083&q=45&auto=format&w=496&fit=clip" /><figcaption><span class="caption">Elements affecting the popularity of songs change over time and should be continuously explored, say data science researchers. </span> <span class="attribution"><span class="source">(Shutterstock)</span></span></figcaption></figure><p>Music is part of our lives in different ways. We listen to it on our commutes and it resounds through shopping centres. Some of us seek live music at concerts, festivals and shows or rely on music to set the tone and mood of our days.</p>
<p>While we might understand the genres or songs we appreciate, it’s not clear precisely why a certain song is more appealing or popular. Perhaps the lyrics speak to an experience? Perhaps the energy makes it appealing? These questions are important to answer for music industry professionals, and <a href="https://theconversation.com/how-data-is-transforming-the-music-industry-70940">analyzing data</a> is a key part of this.</p>
<p>At Carleton University, a group of data science researchers sought to answer the question: “What descriptive features of a song make it popular on music/online platforms?”</p>
<h2>Revenue in the music industry</h2>
<p>Revenue in the music industry <a href="https://doi.org/10.1509/jm.14.0473">is derived from two sources that are affected by different factors: live music and recorded music</a>. During the pandemic, although live music income dropped due to the cancellation of in-person performances, the <a href="https://doi.org/10.1371/journal.pone.0267640">income from streaming</a> rose. </p>
<p>As digital platforms like Spotify and TikTok have grown, <a href="https://doi.org/10.5753/sbcm.2019.10436">the majority of music revenue has come to be contributed by digital media, mostly music streaming</a>. How and whether this <a href="https://theconversation.com/artists-spotify-criticisms-point-to-larger-ways-musicians-lose-with-streaming-heres-3-changes-to-help-in-canada-176526">revenue reaches singers and songwriters at large</a> is another matter. </p>
<h2>Popularity on digital platforms</h2>
<p>The popularity of a song on digital platforms is considered a measure of the revenue the song may generate.</p>
<p>As such, producers seek to answer questions like “<a href="https://doi.org/10.1098/rsos.171274">How can we make the song more popular?</a>” and “<a href="https://doi.org/10.1109/ICMLA.2019.00149">What are the characteristics of songs that make it the top charts?</a>” </p>
<p>With collaborators <a href="https://www.linkedin.com/in/laura-colley/">Laura Colley</a>, <a href="https://www.linkedin.com/in/andrew-dybka/">Andrew Dybka</a>, Adam Gauthier, Jacob Laboissonniere, Alexandre Mougeot and Nayeeb Mowla, we produced a systematic study that collected data from YouTube, Twitter, TikTok, Spotify and Billboard (<a href="https://www.billboard.com/charts/hot-100">Billboard Hot-100</a>, sometimes also denoted by data researchers as “<a href="https://data.world/bigml/association-discovery">Billboard hot top</a>” or in our work and others’ work, “Billboard Top-100”).</p>
<p>We linked the datasets from the different platforms with Spotify’s acoustic descriptive metric or “descriptive features” for songs. These features have been derived <a href="https://www.billboard.com/music/music-news/echo-nest-columbia-university-launch-million-song-dataset-1178990/">from a dataset which yielded categories for measuring and analyzing qualities of songs</a>. Spotify’s <a href="https://www.theguardian.com/technology/2014/mar/06/spotify-echo-nest-streaming-music-deal">metrics capture</a> <a href="https://doi.org/10.1098/rsos.171274">descriptive features such as</a> <em>acousticness</em>, <em>energy</em>, <em>danceability</em> and <em>instrumentalness</em> (the collection of instruments and voices in a given piece). </p>
<p>We sought to find trends and analyze the relationship between songs’ descriptive features and their popularity.</p>
<p>The rankings on the weekly <a href="https://www.billboard.com/charts/hot-100/">Billboard Hot-100</a> are based on sales, online streams and radio plays in the United States.</p>
<p>The analysis we performed by looking at Spotify and Billboard revealed insights that are useful for the music industry.</p>
<figure class="align-center ">
<img alt="A person seen grooving in their car." src="https://images.theconversation.com/files/487408/original/file-20220929-12-z8jn23.jpg?ixlib=rb-1.1.0&q=45&auto=format&w=754&fit=clip" srcset="https://images.theconversation.com/files/487408/original/file-20220929-12-z8jn23.jpg?ixlib=rb-1.1.0&q=45&auto=format&w=600&h=399&fit=crop&dpr=1 600w, https://images.theconversation.com/files/487408/original/file-20220929-12-z8jn23.jpg?ixlib=rb-1.1.0&q=30&auto=format&w=600&h=399&fit=crop&dpr=2 1200w, https://images.theconversation.com/files/487408/original/file-20220929-12-z8jn23.jpg?ixlib=rb-1.1.0&q=15&auto=format&w=600&h=399&fit=crop&dpr=3 1800w, https://images.theconversation.com/files/487408/original/file-20220929-12-z8jn23.jpg?ixlib=rb-1.1.0&q=45&auto=format&w=754&h=502&fit=crop&dpr=1 754w, https://images.theconversation.com/files/487408/original/file-20220929-12-z8jn23.jpg?ixlib=rb-1.1.0&q=30&auto=format&w=754&h=502&fit=crop&dpr=2 1508w, https://images.theconversation.com/files/487408/original/file-20220929-12-z8jn23.jpg?ixlib=rb-1.1.0&q=15&auto=format&w=754&h=502&fit=crop&dpr=3 2262w" sizes="(min-width: 1466px) 754px, (max-width: 599px) 100vw, (min-width: 600px) 600px, 237px">
<figcaption>
<span class="caption">How well can ‘danceability,’ and other factors as rated by Spotify, predict music popularity?</span>
<span class="attribution"><span class="source">(Shutterstock)</span></span>
</figcaption>
</figure>
<h2>What predicts a Billboard hit?</h2>
<p>To perform <a href="https://ieeexplore.ieee.org/document/9842568">this study</a>, we used two different data sets pertaining to songs that <a href="https://www.npr.org/sections/therecord/2013/08/16/207879695/how-the-hot-100-became-americas-hit-barometer">were Billboard hits</a> <a href="https://data.world/kcmillersean/billboard-hot-100-1958-2017">from the early 1940s to 2020</a> and Spotify data related to over 600,000 tracks and over one million artists.</p>
<p>Interestingly, we found no substantial correlations between the number of weeks a song remained on the charts, as a measure of popularity, and the acoustic features included in the study. </p>
<p>Our analysis determined that newer songs tend to last longer on the charts and that a song’s popularity affects how long it stays on the charts. </p>
<p>In a related study, researchers collected data for Billboard’s Hot 100 from 1958 to 2013 and found that <a href="https://doi.org/10.1007/978-3-319-13734-6_36">songs with a higher <em>tempo</em> and <em>danceability</em> often get a higher peak position on the Billboard charts</a>. </p>
<h2>Predicting Spotify song popularity</h2>
<p>We also used the songs’ features to generate machine learning models to predict Spotify song popularity. Preliminary results concluded that features are not linearly correlated, with some expected exceptions including songs’ <em>energy</em>. </p>
<p>This indicated that the Spotify metrics we studied — including <em>acousticness</em>, <em>danceability</em>, <em>duration</em>, <em>energy</em>, <em>explicitness</em>, <em>instrumentalness</em>, <em>liveness</em>, <em>speechiness</em> (a measure of the presence of spoken words in a song), <em>tempo</em> and <em>release year</em> — were not strong predictors of the song’s popularity.</p>
<p>The majority of songs in the Spotify dataset were not listed as explicit, tended to have low <em>instrumentalness</em> and <em>speechiness</em>, and were typically recent songs. </p>
<p>Although one may think that some features that are innate to certain songs make them more popular, our study revealed that popularity can not be attributed solely to quantifiable acoustic elements. </p>
<p>This means that song makers and consumers must consider other contextual factors beyond the musical features, as captured by Spotify’s measurables, that may contribute to the song’s success. </p>
<h2>Elements affecting popularity shift</h2>
<p>Our study reinforces that elements affecting the popularity of songs change over time and should be continuously explored. </p>
<p>For example, <a href="https://doi.org/10.1098%2Frsos.171274">in songs produced between 1985 and 2015 in the United Kingdom, songs produced by female artists were more successful</a>. </p>
<figure class="align-left ">
<img alt="Woman holding a microphone seen flexing her bicep." src="https://images.theconversation.com/files/487411/original/file-20220929-25-g603nl.jpg?ixlib=rb-1.1.0&q=45&auto=format&w=237&fit=clip" srcset="https://images.theconversation.com/files/487411/original/file-20220929-25-g603nl.jpg?ixlib=rb-1.1.0&q=45&auto=format&w=600&h=422&fit=crop&dpr=1 600w, https://images.theconversation.com/files/487411/original/file-20220929-25-g603nl.jpg?ixlib=rb-1.1.0&q=30&auto=format&w=600&h=422&fit=crop&dpr=2 1200w, https://images.theconversation.com/files/487411/original/file-20220929-25-g603nl.jpg?ixlib=rb-1.1.0&q=15&auto=format&w=600&h=422&fit=crop&dpr=3 1800w, https://images.theconversation.com/files/487411/original/file-20220929-25-g603nl.jpg?ixlib=rb-1.1.0&q=45&auto=format&w=754&h=530&fit=crop&dpr=1 754w, https://images.theconversation.com/files/487411/original/file-20220929-25-g603nl.jpg?ixlib=rb-1.1.0&q=30&auto=format&w=754&h=530&fit=crop&dpr=2 1508w, https://images.theconversation.com/files/487411/original/file-20220929-25-g603nl.jpg?ixlib=rb-1.1.0&q=15&auto=format&w=754&h=530&fit=crop&dpr=3 2262w" sizes="(min-width: 1466px) 754px, (max-width: 599px) 100vw, (min-width: 600px) 600px, 237px">
<figcaption>
<span class="caption">Melanie Chisholm (Sporty Spice) of the Spice Girls seen in December 2007.</span>
<span class="attribution"><span class="source">THE CANADIAN PRESS/Jonathan Hayward</span></span>
</figcaption>
</figure>
<p>Other aspects may substantially contribute to the success of a song. Data scientists have proposed <a href="https://doi.org/10.1371/journal.pone.0244576">simplicity of the lyrics</a>, the advertising and <a href="https://www.ipr.edu/blogs/audio-production/what-are-the-elements-of-popular-music/">distribution plans</a> as potential predictors of songs’ popularity. </p>
<hr>
<p>
<em>
<strong>
Read more:
<a href="https://theconversation.com/beatles-get-back-documentary-reveals-how-creativity-doesnt-happen-on-its-own-182380">Beatles 'Get Back' documentary reveals how creativity doesn't happen on its own</a>
</strong>
</em>
</p>
<hr>
<h2>Attached listeners</h2>
<p>Many musicians and producers make use of popular events and marketing strategies to advertise songs. Such events create social engagements and <a href="https://doi.org/10.3389/fpsyg.2018.02682">audience involvement</a> which attaches the listener to the song being performed. </p>
<p>For the public, <a href="https://www.osheaga.com/en">live music events</a>, following long lockdowns, have been opportune for reuniting friends, and <a href="https://ottawabluesfest.ca">enjoying live artistry and</a> entertainment.</p>
<p>While attending a music event or listening to a song, we invite you to reflect on what it is about the song that makes you enjoy it.</p><img src="https://counter.theconversation.com/content/189052/count.gif" alt="The Conversation" width="1" height="1" />
<p class="fine-print"><em><span>The authors do not work for, consult, own shares in or receive funding from any company or organisation that would benefit from this article, and have disclosed no relevant affiliations beyond their academic appointment.</span></em></p>Whether we’re talking songs that are popular on Spotify, or were Billboard hits through the ‘40s up to recent years, popularity cannot be attributed solely to quantifiable acoustic elements.Hoda Khalil, Postdoctoral Research Associate and Contract Instructor, Systems and Computer Engineering, Carleton UniversityGabriel Wainer, Professor, Systems and Computer Engineering, Carleton UniversityKevin Dick, Postdoctoral Fellow in the Department of Systems and Computer Engineering, Carleton UniversityLicensed as Creative Commons – attribution, no derivatives.tag:theconversation.com,2011:article/1872742022-07-20T20:08:52Z2022-07-20T20:08:52ZWhat do TikTok, Bunnings, eBay and Netflix have in common? They’re all hyper-collectors<figure><img src="https://images.theconversation.com/files/474987/original/file-20220719-6978-2qdmfk.jpg?ixlib=rb-1.1.0&q=45&auto=format&w=496&fit=clip" /><figcaption><span class="caption">
</span> <span class="attribution"><span class="source">Shutterstock</span></span></figcaption></figure><p>You walk into a shopping centre to buy some groceries. Without your knowledge, an electronic scan of your face is taken by in-store surveillance cameras and stored in an online database. Each time you return to that store, your “faceprint” is compared with those of people wanted for shoplifting or violence.</p>
<p>This might sound like science fiction but it’s the reality for many of us. By failing to take our digital privacy seriously – as former human rights commissioner Ed Santow has warned – Australia is “<a href="https://www.theage.com.au/national/we-must-not-sleepwalk-into-mass-surveillance-20220630-p5ay0q.html">sleepwalking</a>” its way into mass surveillance.</p>
<h2>Privacy and the digital environment</h2>
<p>Of course, companies have been collecting personal information for decades. If you’ve ever signed up to a loyalty program like FlyBuys then you’ve performed what marketing agencies call a “<a href="https://www.choice.com.au/consumers-and-data/data-collection-and-use/who-has-your-data/articles/loyalty-program-data-collection">value exchange</a>”. In return for benefits from the company (like discounted prices or special offers), you’ve handed over details of who you are, what you buy, and how often you buy it.</p>
<p>Consumer data is big business. In 2019, a <a href="https://www.webfx.com/blog/internet/what-are-data-brokers-and-what-is-your-data-worth-infographic/">report</a> from digital marketers WebFX showed that data from around 1,400 loyalty programs was routinely being traded across the globe as part of an industry <a href="https://clearcode.cc/blog/what-is-data-broker/">worth around US$200 billion</a>. That same year, the Australian Competition and Consumer Commission’s <a href="https://www.accc.gov.au/publications/customer-loyalty-schemes-final-report">review of loyalty schemes</a> revealed how many of these loyalty schemes lacked data transparency and even discriminated against vulnerable customers.</p>
<p>But the digital environment is making data collection even easier. When you <a href="https://onlinemasters.ohio.edu/blog/netflix-data/">watch Netflix</a>, for example, the company knows what you watch, when you watch it, and how long you watch it for. But they go further, also <a href="https://seleritysas.com/blog/2019/04/05/how-netflix-used-big-data-and-analytics-to-generate-billions/">capturing data</a> on which scenes or episodes you watch repeatedly, the ratings of your content, the number of searches you perform and what you search for.</p>
<h2>Hyper-collection: a new challenge to privacy</h2>
<p>Late last year, the controversial tech company ClearView AI was <a href="https://www.oaic.gov.au/updates/news-and-media/clearview-ai-breached-australians-privacy">ordered</a> by the Australian information commissioner to stop “scraping” social media for the pictures it was collecting in its massive facial recognition database. Just this month, the commissioner was investigating several retailers for <a href="https://www.abc.net.au/news/2022-07-13/bunnings-kmart-investigated-over-facial-recognition-technology/101233372">creating facial profiles</a> of the customers in their stores.</p>
<p>This new phenomenon – “hyper-collection” – represents a growing trend by large companies to collect, sort, analyse and use more information than they need, usually in covert or passive ways. In many cases, hyper-collection is not supported by a truly legitimate commercial or legal purpose.</p>
<h2>Digital privacy laws and hyper-collection</h2>
<p>Hyper-collection is a major problem in Australia for three reasons.</p>
<p>First, Australia’s privacy law wasn’t prepared for the likes of Netflix and TikTok. Despite <a href="https://www.oaic.gov.au/privacy/the-privacy-act/history-of-the-privacy-act">numerous amendments</a>, the <a href="https://www.oaic.gov.au/privacy/the-privacy-act">Privacy Act</a> dates back to the late 1980s. Although former Attorney-General Christian Porter <a href="https://www.ag.gov.au/integrity/consultations/review-privacy-act-1988">announced a review</a> of the Act in late 2019, it has been held up by the recent change of government.</p>
<p>Second, Australian privacy laws are unlikely on their own to threaten the profit base of foreign companies, especially those located in China. The Information Commissioner has the power to order companies to take certain actions – like it <a href="https://www.afr.com/policy/foreign-affairs/australia-s-tiktok-data-vulnerable-to-access-by-china-staff-20220712-p5b10f">did with Uber in 2021</a> – and can enforce these through court orders. But the penalties aren’t really big enough to discourage companies with profits in the billions of dollars.</p>
<hr>
<p>
<em>
<strong>
Read more:
<a href="https://theconversation.com/83-of-australians-want-tougher-privacy-laws-nows-your-chance-to-tell-the-government-what-you-want-149535">83% of Australians want tougher privacy laws. Now’s your chance to tell the government what you want</a>
</strong>
</em>
</p>
<hr>
<p>Third, hyper-collection is often enabled by the vague consents we give to get access to the services these companies provide. Bunnings, for example, argued that its collection of your faceprint was allowed because <a href="https://ia.acs.org.au/article/2022/bunnings-doubles-down-on-facial-recognition.html">signs at the entry to their stores</a> told customers facial recognition might be used. Online marketplaces like eBay, Amazon, Kogan and Catch, meanwhile, supply “<a href="https://www.accc.gov.au/media-release/concerning-issues-for-consumers-and-sellers-on-online-marketplaces">bundled consents</a>” – basically, you have to consent to their privacy policies as a condition of using their services. No consent, no access.</p>
<h2>TikTok and hyper-collection</h2>
<p>TikTok (owned by Chinese company ByteDance) has largely replaced YouTube as a way of creating and sharing online videos. The app is powered by an algorithm has already drawn <a href="https://theconversation.com/tiktoks-secret-algorithm-is-its-greatest-strength-and-could-also-be-its-undoing-176605">criticism</a> for routinely collecting data about users, as well as the ByteDance’s secretive approach to <a href="https://www.lowyinstitute.org/the-interpreter/unique-power-tiktok-s-algorithm">content moderation and censorship</a>.</p>
<p>For years, TikTok executives have been telling governments that <a href="https://www.aspistrategist.org.au/its-time-tiktok-australia-came-clean/">data isn’t stored in servers on the Chinese mainland</a>. But these promises might be hollow in the wake of recent allegations.</p>
<p>Cybersecurity experts now claim that not only does the TikTok app <a href="https://www.smartcompany.com.au/technology/tiktok-chinese-servers-aussie-cybersecurity/">routinely connect to Chinese servers</a>, but that users’ data is accessible by ByteDance employees, including the mysterious Beijing-based “Master Admin”, which has <a href="https://www.buzzfeednews.com/article/emilybakerwhite/tiktok-tapes-us-user-data-china-bytedance-access">access to every user’s personal information</a>.</p>
<p>Then, just this week, it was alleged that TikTok (owned by Chinese company ByteDance) can also access <a href="https://www.abc.net.au/news/2022-07-18/tiktok-users-warned-the-platform-is-harvesting-personal-data/13977370">almost all the data</a> contained on the phone it is installed on – including photos, calendars and emails.</p>
<p>Under China’s national security laws, the government can order tech companies to <a href="https://www.sbs.com.au/news/article/so-what-if-china-can-access-your-tiktok-data/mr1anx97k">pass on that information</a> to police or intelligence agencies.</p>
<h2>What options do we have?</h2>
<p>Unlike a physical store, we don’t get a lot of choice about consenting to digital companies’ privacy policies and how they collect our information.</p>
<p>One option – supported by encryption expert Vanessa Teague at ANU – is for consumers simply to delete offending apps until their creators are <a href="https://www.sbs.com.au/news/article/so-what-if-china-can-access-your-tiktok-data/mr1anx97k">willing to submit to greater data transparency</a>. Of course, this means locking ourselves out of those services, and it will only have a big impact in the company if enough Australians join in.</p>
<hr>
<p>
<em>
<strong>
Read more:
<a href="https://theconversation.com/facial-recognition-is-on-the-rise-but-the-law-is-lagging-a-long-way-behind-185510">Facial recognition is on the rise – but the law is lagging a long way behind</a>
</strong>
</em>
</p>
<hr>
<p>Another option is “opting-out” of intrusive data collection. We’ve done this before – when My Health records became mandatory in 2019, a record number of us <a href="https://www.yourlifechoices.com.au/health/my-health-record-an-expensive-white-elephant-critics-say/">opted out</a>. Though these opt-outs reduced the usefulness of that <a href="https://www.theguardian.com/commentisfree/2018/jul/20/there-is-no-social-license-for-my-health-record-australians-should-reject-it">digital health record program</a>, they did demonstrate that Australians can take their data privacy seriously. </p>
<p>But how exactly can Australians opt-out of a massive social app like TikTok? Right now, they can’t – perhaps the government needs to explore a solution as part of its review.</p>
<p>A further option being explored by the Privacy Act review is whether to create new laws that would allow individuals to <a href="https://www.ag.gov.au/system/files/2020-10/privacy-act-review-terms-of-reference.pdf">sue companies for damages for breaches of privacy</a>. While lawsuits are expensive and time-consuming, they might just deliver the kind of financial damage to big companies that could change their behaviour.</p>
<p>No matter which option we take, Australians need to start getting more savvy with their data privacy. This might just mean we actually read those terms and conditions before agreeing, and being prepared to “vote with our feet” if companies won’t be honest about what they’re doing with our personal information.</p><img src="https://counter.theconversation.com/content/187274/count.gif" alt="The Conversation" width="1" height="1" />
<p class="fine-print"><em><span>Brendan Walker-Munro receives funding from the Australian Government through Trusted Autonomous Systems, a Defence Cooperative Research Centre funded through the Next Generation Technologies Fund. </span></em></p>Australians – and Australian governments – need to get more savvy about data privacyBrendan Walker-Munro, Senior Research Fellow, The University of QueenslandLicensed as Creative Commons – attribution, no derivatives.tag:theconversation.com,2011:article/1845312022-06-27T17:15:50Z2022-06-27T17:15:50ZThe dangers of big data extend to farming<figure><img src="https://images.theconversation.com/files/470855/original/file-20220624-19-jc1m5e.jpg?ixlib=rb-1.1.0&rect=54%2C46%2C5121%2C3391&q=45&auto=format&w=496&fit=clip" /><figcaption><span class="caption">Smart technology is demonstrated on a farm in Newark, Mo.</span> <span class="attribution"><span class="source">(Dilip Vishwanat/AP Images/U.S. Cellular)</span></span></figcaption></figure><iframe style="width: 100%; height: 100px; border: none; position: relative; z-index: 1;" allowtransparency="" allow="clipboard-read; clipboard-write" src="https://narrations.ad-auris.com/widget/the-conversation-canada/the-dangers-of-big-data-extend-to-farming" width="100%" height="400"></iframe>
<p>Most internet users are by now aware of the vulnerability of their personal data. When the <a href="https://www.nytimes.com/2018/04/04/us/politics/cambridge-analytica-scandal-fallout.html">news broke that tech companies misuse and manipulate our personal data</a>, there was a widespread <a href="https://www.economist.com/briefing/2018/01/20/the-techlash-against-amazon-facebook-and-google-and-what-they-can-do">“techlash” against the corporate giants Facebook, Amazon and Google</a>.</p>
<p>The explicit <a href="https://www.theguardian.com/books/2019/oct/04/shoshana-zuboff-surveillance-capitalism-assault-human-automomy-digital-privacy">motive for data harvesting</a> is the prediction of consumer wants and needs. And scholars and activists have spent years exposing the dangerous effects of big data practices on individual <a href="https://us.macmillan.com/books/9781250074317/automatinginequality">privacy</a> and civil liberties. </p>
<p>Researchers have also shown how biased big data and opaque artificial intelligence have reproduced racism, classism and many forms of <a href="https://nyupress.org/9781479837243/algorithms-of-oppression/">inequity</a>. </p>
<p>However less attention has been paid to agriculture firms such as Monsanto or John Deere. And yet, they are some of the largest, concentrated corporations in North America who increasingly centre their profits on the collection, processing and sale of big data.</p>
<p>We should pay close attention. Agricultural big data are likely to have far-reaching detrimental environmental and social impacts. </p>
<h2>Maximizing profits: The past as a road map</h2>
<p>While farmers have used satellite weather data for decades, today’s big data practices are different. Data get collected by sensors built into farm machinery, called “precision” farm equipment. Then, corporate data scientists aggregate and “mine” these data for insights. </p>
<p>While there are benefits to farmers who buy the data-driven insights, there are clear and potentially incommensurate benefits for the corporations. Both the data-driven insights as well as the datasets are sources of profit for private firms. </p>
<p>We can infer what agribusinesses might do with big data based on insights from other sectors but also by how agribusinesses have acted in the past. </p>
<p>It is likely that firms collecting data from farms about weather and pest pressures will be able to predict which products are most needed where and then use this information to maximize profit. </p>
<p>Companies supplying farmers with seeds and chemicals have for years used <a href="https://www.wsj.com/articles/SB10001424127887323777204578189391813881534">price discrimination</a>, selectively setting higher prices for inputs within those demographics or regions which are seen to depend on them. </p>
<p>Big agricultural data could entrench the market advantage of large agribusinesses.</p>
<h2>Agricultural big data</h2>
<p>Companies collecting and controlling the most data are likely to accrue the most power. This is because of the scale-driven proposition of artificial intelligence: AI is only as good as the data which <a href="https://www.nytimes.com/2016/10/17/technology/ibm-is-counting-on-its-bet-on-watson-and-paying-big-money-for-it.html">feeds</a> it. </p>
<p>We have seen this advantage play out with Google and Amazon, who were first out of the gate collecting internet data. Their lead has enabled the development of robust AI systems like Google Search and Alexa. </p>
<p>That the powerful can become even more powerful in the big data era is relevant for agriculture where the sector has long been <a href="https://philhoward.net/2020/09/27/the-state-of-concentration-in-global-food-and-agriculture-industries/">dominated</a> by only a handful of companies. </p>
<h2>A marriage made in hell</h2>
<p>In 2018, Monsanto and Bayer, the two biggest agribusinesses, merged in a transaction worth US$63 billion. At the time, Monsanto owned over 80 per cent of all the genetically modified seeds in the world. Unprecedented in its scope, the merger was described as “marriage made in hell” given its <a href="https://www.theguardian.com/business/2016/sep/14/bayer-takeover-monsanto-66-billion-deal">anti-competition ramifications</a>. </p>
<p>The side of the story that’s relevant to big data is that new Bayer has the capability to access data from <a href="https://cban.ca/gmos/issues/monsanto/">almost half of all farmers in North America</a>.</p>
<p>The ins and outs of Bayer’s AI system are impossible to assess because its workings are protected from scrutiny by trade secrecy law, rendering it a “<a href="https://www.hup.harvard.edu/catalog.php?isbn=9780674970847">pernicious black box</a>.”</p>
<figure class="align-center ">
<img alt="Pumpjacks draw oil out of the ground as a deer stands in a yellow canola field" src="https://images.theconversation.com/files/467859/original/file-20220608-24-bf825l.JPG?ixlib=rb-1.1.0&rect=125%2C59%2C3868%2C2185&q=45&auto=format&w=754&fit=clip" srcset="https://images.theconversation.com/files/467859/original/file-20220608-24-bf825l.JPG?ixlib=rb-1.1.0&q=45&auto=format&w=600&h=338&fit=crop&dpr=1 600w, https://images.theconversation.com/files/467859/original/file-20220608-24-bf825l.JPG?ixlib=rb-1.1.0&q=30&auto=format&w=600&h=338&fit=crop&dpr=2 1200w, https://images.theconversation.com/files/467859/original/file-20220608-24-bf825l.JPG?ixlib=rb-1.1.0&q=15&auto=format&w=600&h=338&fit=crop&dpr=3 1800w, https://images.theconversation.com/files/467859/original/file-20220608-24-bf825l.JPG?ixlib=rb-1.1.0&q=45&auto=format&w=754&h=424&fit=crop&dpr=1 754w, https://images.theconversation.com/files/467859/original/file-20220608-24-bf825l.JPG?ixlib=rb-1.1.0&q=30&auto=format&w=754&h=424&fit=crop&dpr=2 1508w, https://images.theconversation.com/files/467859/original/file-20220608-24-bf825l.JPG?ixlib=rb-1.1.0&q=15&auto=format&w=754&h=424&fit=crop&dpr=3 2262w" sizes="(min-width: 1466px) 754px, (max-width: 599px) 100vw, (min-width: 600px) 600px, 237px">
<figcaption>
<span class="caption">In just a few decades, canola has become one of the world’s most important oilseed crops and the most profitable commodity for Canadian farmers.</span>
<span class="attribution"><span class="source">THE CANADIAN PRESS/Jeff McIntosh</span></span>
</figcaption>
</figure>
<p>As with Facebook, Google and Amazon, neither the firms’ datasets or the AI processes by which they translate data into advice are transparent. Corporations justify tight data control as a means to protect consumers from privacy breaches and actions of nefarious hackers. </p>
<p>But as with social media companies, there is no democratic institutional oversight over agricultural big data. Most data breaches have so far been unearthed by <a href="https://www.propublica.org/article/machine-bias-risk-assessments-in-criminal-sentencing">watchdog sleuthing</a>. Misuses of agricultural data will likely be worked out, as they have been for internet data — in the court of public opinion. </p>
<h2>Biased data sets</h2>
<p>Even without access to commercial code or data sets, one can analyze what is happening at the surface of agricultural big data — what media studies scholar Ian Bogost calls “<a href="https://mitpress.mit.edu/books/persuasive-games">procedural rhetoric</a>.” </p>
<p>Many folks claim that big data and AI are efficient, objective, valuable and all-powerful. But contrary to the view that data are “raw” and “drive” advice (a fallacious view I call the <a href="https://www.mqup.ca/immaculate-conception-of-data--the-products-9780228011224.php"><em>Immaculate Conception of Data</em></a>), people define and make relevant the range and content of big data and AI categories. Computer scientists and engineers design algorithms and build datasets. </p>
<p>For example, company scientists design the dominant commercial agricultural AI to include data only on a small selection of major agronomic commodity crops, those grown on large farms. </p>
<p>This <a href="https://doi.org/10.1016/j.njas.2019.03.001">design bias</a> toward large, commodity crop and capital-intensive farms presents implications at a large scale, such that our food system could become increasingly characterized by an industrial mode of agriculture. </p>
<h2>Big data is not the solution to food problems</h2>
<p><a href="https://theconversation.com/5-technologies-that-will-help-make-the-food-system-carbon-neutral-182846">Many scholars</a> believe data-driven readings of the world are a primary means of solving food system problems. </p>
<p>Indeed, the UN organization Global Open Data for Agriculture and Nutrition and others <a href="https://theconversation.com/african-agriculture-has-a-lot-to-gain-from-increased-access-to-big-data-101400">call for opening agricultural data</a>, as if data access was synonymous with broader social and environmental justice. </p>
<p>Industrial agriculture has <a href="https://www.ipes-food.org/_img/upload/files/LFMExecSummaryEN.pdf">proven negative impacts</a> including on global <a href="https://doi.org/10.1038/s43016-021-00225-9">climate change</a>. There is growing consensus that we need to diversify our food system at all levels to foster a <a href="https://doi.org/10.1002/fes3.154">sustainability transition</a> in agriculture.</p>
<p>Beyond data access and infrastructure, digital democracy calls for a fundamental redistribution of decision-making power from a small number of corporate stakeholders to a wider group of citizens who can help answer these questions: What kind of food system do we want? Which farming techniques and technologies will help us get there?</p><img src="https://counter.theconversation.com/content/184531/count.gif" alt="The Conversation" width="1" height="1" />
<p class="fine-print"><em><span>Kelly Bronson receives funding from the Social Sciences and Humanities Research Council</span></em></p>Big data from social media have been revealed as biased, but we should also pay attention to agriculture firms whose play for big data is likely to have detrimental environmental and social impacts.Kelly Bronson, Canada Research Chair in Science & Society, L’Université d’Ottawa/University of OttawaLicensed as Creative Commons – attribution, no derivatives.tag:theconversation.com,2011:article/1850382022-06-19T19:53:40Z2022-06-19T19:53:40ZInsurance firms can skim your online data to price your insurance — and there’s little in the law to stop this<figure><img src="https://images.theconversation.com/files/469391/original/file-20220617-24-txo2j0.jpeg?ixlib=rb-1.1.0&rect=58%2C69%2C7684%2C5084&q=45&auto=format&w=496&fit=clip" /><figcaption><span class="caption">
</span> <span class="attribution"><span class="source">Shutterstock</span></span></figcaption></figure><p>What if your insurer was tracking your online data to price your car insurance? Seems far-fetched, right? </p>
<p>Yet there is predictive value in the digital traces we leave online. And insurers may use data collection and analytics tools to find our data and use it to price insurance services. </p>
<p>For instance, <a href="https://pubmed.ncbi.nlm.nih.gov/27849366/">some</a> <a href="https://www.researchgate.net/publication/350525424_Smartphone_Operating_System_Preference_Based_On_Different_Personality_Lifestyle_Traits_Of_The_Consumer">studies</a> <a href="https://www.nber.org/papers/w24771#fromrss">have</a> found a correlation between whether an individual uses an Apple or Android phone and their likelihood of exhibiting certain personality traits. </p>
<p>In one example, US insurance broker Jerry analysed the driving behaviour of some 20,000 people to conclude Android users are <a href="https://getjerry.com/studies/sorry-iphone-fans-android-users-are-safer-drivers">safer drivers</a> than iPhone users. What’s stopping insurers from referring to such reports to price their insurance?</p>
<p>Our latest <a href="https://www.sciencedirect.com/science/article/abs/pii/S0267364922000152">research</a> shows Australian consumers have no real control over how data about them, and posted by them, might be collected and used by insurers. </p>
<p>Looking at several examples from customer loyalty schemes and social media, we found insurers can access vast amounts of consumer data under Australia’s <a href="https://www.ag.gov.au/integrity/consultations/review-privacy-act-1988">weak privacy laws</a>. </p>
<figure class="align-center zoomable">
<a href="https://images.theconversation.com/files/469394/original/file-20220617-21-k84rhp.jpeg?ixlib=rb-1.1.0&q=45&auto=format&w=1000&fit=clip"><img alt="A person's hands are visible holding an Apple phone on the left (screen facing forward), and a generic Android on the right." src="https://images.theconversation.com/files/469394/original/file-20220617-21-k84rhp.jpeg?ixlib=rb-1.1.0&q=45&auto=format&w=754&fit=clip" srcset="https://images.theconversation.com/files/469394/original/file-20220617-21-k84rhp.jpeg?ixlib=rb-1.1.0&q=45&auto=format&w=600&h=450&fit=crop&dpr=1 600w, https://images.theconversation.com/files/469394/original/file-20220617-21-k84rhp.jpeg?ixlib=rb-1.1.0&q=30&auto=format&w=600&h=450&fit=crop&dpr=2 1200w, https://images.theconversation.com/files/469394/original/file-20220617-21-k84rhp.jpeg?ixlib=rb-1.1.0&q=15&auto=format&w=600&h=450&fit=crop&dpr=3 1800w, https://images.theconversation.com/files/469394/original/file-20220617-21-k84rhp.jpeg?ixlib=rb-1.1.0&q=45&auto=format&w=754&h=566&fit=crop&dpr=1 754w, https://images.theconversation.com/files/469394/original/file-20220617-21-k84rhp.jpeg?ixlib=rb-1.1.0&q=30&auto=format&w=754&h=566&fit=crop&dpr=2 1508w, https://images.theconversation.com/files/469394/original/file-20220617-21-k84rhp.jpeg?ixlib=rb-1.1.0&q=15&auto=format&w=754&h=566&fit=crop&dpr=3 2262w" sizes="(min-width: 1466px) 754px, (max-width: 599px) 100vw, (min-width: 600px) 600px, 237px"></a>
<figcaption>
<span class="caption">How would you feel if a detail as menial as the brand of your phone was used to price your car insurance?</span>
<span class="attribution"><span class="source">Shutterstock</span></span>
</figcaption>
</figure>
<h2>Your data is already out there</h2>
<p>Insurers are already using big data to price consumer insurance through personalised pricing, according to evidence gathered by industry regulators in the <a href="https://www.fca.org.uk/publication/feedback/fs16-05.pdf">United Kingdom</a>, <a href="https://register.eiopa.europa.eu/Publications/EIOPA_BigDataAnalytics_ThematicReview_April2019.pdf">European Union</a> and <a href="https://www.dfs.ny.gov/industry_guidance/circular_letters/cl2019_01">United States</a>.</p>
<p>Consumers often “agree” to all kinds of data collection and privacy policies, such as those used in loyalty schemes (who doesn’t like freebies?) and by social media companies. But they have no control over how their data are used once it’s handed over.</p>
<p>There are far-reaching inferences that can be drawn from data collected through loyalty programs and social media platforms – and these may be uncomfortable, or even highly sensitive.</p>
<p>Researchers using data analytics and machine learning have claimed to build models that can guess a person’s sexual orientation from pictures of <a href="https://osf.io/zn79k/">their face</a>, or their suicidal tendencies from <a href="https://www.sciencedirect.com/science/article/pii/S2214782915000160">posts on Twitter</a>.</p>
<p>Think about all the details revealed from a grocery shopping history alone: diet, household size, addictions, health conditions and social background, among others. In the case of social media, a user’s posts, pictures, likes, and links to various groups can be used to draw a precise picture of that individual.</p>
<p>What’s more is Australia has a <a href="https://www.cdr.gov.au/">Consumer Data Right</a> which already requires banks to share consumers’ banking data (at the consumer’s request) with another bank or app, such as to access a new service or offer. </p>
<p>The regime is actively being expanded to other parts of the economy including the energy sector, with the idea being competitors could use information on energy usage to make competitive offers. </p>
<p>The Consumer Data Right is advertised as <a href="https://www.cdr.gov.au/">empowering</a> for consumers – enabling access to new services and offers, and providing people with choice, convenience and control over their data. </p>
<p>In practice, however, it means insurance firms accredited under the program can require you to share your banking data in exchange for insurance services.</p>
<p>The previous Coalition government also <a href="https://ministers.treasury.gov.au/ministers/jane-hume-2020/media-releases/more-power-compare-and-switch-telco-providers-and-share">proposed “open finance”</a>, which would expand the Consumer Data Right to include access to your insurance and superannuation data. This hasn’t happened yet, but it’s likely the new Albanese government will look into it.</p>
<hr>
<p>
<em>
<strong>
Read more:
<a href="https://theconversation.com/soft-terms-like-open-and-sharing-dont-tell-the-true-story-of-your-data-95521">Soft terms like 'open' and 'sharing' don't tell the true story of your data</a>
</strong>
</em>
</p>
<hr>
<h2>Why more data in insurers’ hands may be bad news</h2>
<p>There are plenty of reasons to be concerned about insurers collecting and using increasingly detailed data about people for insurance pricing and claims management. </p>
<p>For one, large-scale data collection provides incentives for cyber attacks. Even if data is held in anonymised form, it can be <a href="https://techcrunch.com/2019/07/24/researchers-spotlight-the-lie-of-anonymous-data/">re-identified</a> with the right tools. </p>
<p>Also, insurers may be able to infer (or at least think they can infer) facts about an individual which they want to keep private, such as their sexual orientation, <a href="https://www.forbes.com/sites/kashmirhill/2012/02/16/how-target-figured-out-a-teen-girl-was-pregnant-before-her-father-did/?sh=794d21176668">pregnancy</a> status or religious beliefs. </p>
<p>There’s plenty of evidence the outputs of artificial intelligence tools employed in mass data analytics can be inaccurate and discriminatory. Insurers’ decisions may then be based on misleading or untrue data. And these tools are so complex it’s often difficult to work out if, or where, errors or bias are present.</p>
<figure class="align-center zoomable">
<a href="https://images.theconversation.com/files/469396/original/file-20220617-13-58ptct.jpeg?ixlib=rb-1.1.0&q=45&auto=format&w=1000&fit=clip"><img alt="A magnifying glass hovers over a Facebook post's likes" src="https://images.theconversation.com/files/469396/original/file-20220617-13-58ptct.jpeg?ixlib=rb-1.1.0&q=45&auto=format&w=754&fit=clip" srcset="https://images.theconversation.com/files/469396/original/file-20220617-13-58ptct.jpeg?ixlib=rb-1.1.0&q=45&auto=format&w=600&h=400&fit=crop&dpr=1 600w, https://images.theconversation.com/files/469396/original/file-20220617-13-58ptct.jpeg?ixlib=rb-1.1.0&q=30&auto=format&w=600&h=400&fit=crop&dpr=2 1200w, https://images.theconversation.com/files/469396/original/file-20220617-13-58ptct.jpeg?ixlib=rb-1.1.0&q=15&auto=format&w=600&h=400&fit=crop&dpr=3 1800w, https://images.theconversation.com/files/469396/original/file-20220617-13-58ptct.jpeg?ixlib=rb-1.1.0&q=45&auto=format&w=754&h=503&fit=crop&dpr=1 754w, https://images.theconversation.com/files/469396/original/file-20220617-13-58ptct.jpeg?ixlib=rb-1.1.0&q=30&auto=format&w=754&h=503&fit=crop&dpr=2 1508w, https://images.theconversation.com/files/469396/original/file-20220617-13-58ptct.jpeg?ixlib=rb-1.1.0&q=15&auto=format&w=754&h=503&fit=crop&dpr=3 2262w" sizes="(min-width: 1466px) 754px, (max-width: 599px) 100vw, (min-width: 600px) 600px, 237px"></a>
<figcaption>
<span class="caption">Each day, people post personal information online. And much of it can be easily accessed by others.</span>
<span class="attribution"><span class="source">Shutterstock</span></span>
</figcaption>
</figure>
<p>Although insurers are meant to pool risk and compensate the unlucky, some might use data to only offer affordable insurance to very low-risk people. Vulnerable consumers may face <a href="https://actuaries.logicaldoc.cloud/download-ticket?ticketId=09c77750-aa90-4ba9-835e-280ae347487b">exclusion</a>. </p>
<p>A more widespread use of data, especially via the Consumer Data Right, will especially disadvantage those who are unable or unwilling to share data with insurers. These people may be low risk, but if they can’t or won’t prove this, they’ll have to pay more than a fair price for their insurance cover. </p>
<p>They may even pay more than what they would have in a pre-Consumer Data Right world. So insurance may move <em>further</em> from a fair price when more personal data are available to insurance firms. </p>
<h2>We need immediate action</h2>
<p>Our <a href="http://www5.austlii.edu.au/au/journals/SydLawRw/2021/20.html">previous research</a> demonstrated that apart from anti-discrimination laws, there are inadequate constraints on how insurers are allowed to use consumers’ data, such as those taken from online sources. </p>
<p>The more insurers base their assessments on data a consumer didn’t directly provide, the harder it will be for that person to understand how their “riskiness” is being assessed. If an insurer requests your transaction history from the last five years, would you know what they are looking for? Such problems will be exacerbated by the expansion of the Consumer Data Right.</p>
<p>Interestingly, insurance firms themselves might <a href="https://www.nature.com/news/can-we-open-the-black-box-of-ai-1.20731">not know</a> how collected data translates into risk for a specific consumer. If their approach is to simply feed data into a complex and opaque artificial intelligence system, all they’ll know is they’re getting a supposedly “better” risk assessment with more data.</p>
<p>Recent <a href="https://theconversation.com/bunnings-kmart-and-the-good-guys-say-they-use-facial-recognition-for-loss-prevention-an-expert-explains-what-it-might-mean-for-you-185126">reports</a> of retailers collecting shopper data for facial recognition have highlighted how important it is for the Albanese government to urgently reform <a href="https://www.ag.gov.au/integrity/consultations/review-privacy-act-1988">our privacy laws</a>, and take a close look at other data laws, including proposals to <a href="https://treasury.gov.au/review/statutory-review-consumer-data-right">expand the Consumer Data Right</a>.</p>
<hr>
<p>
<em>
<strong>
Read more:
<a href="https://theconversation.com/hidden-costs-manipulation-forced-continuity-report-reveals-how-australian-consumers-are-being-duped-online-184450">Hidden costs, manipulation, forced continuity: report reveals how Australian consumers are being duped online</a>
</strong>
</em>
</p>
<hr>
<img src="https://counter.theconversation.com/content/185038/count.gif" alt="The Conversation" width="1" height="1" />
<p class="fine-print"><em><span>Zofia Bednarz receives funding from the Australian Research Council Centre of Excellence on Automated Decision-Making and Society. </span></em></p><p class="fine-print"><em><span>Kayleen Manwaring receives funding from the UNSW Allens Hub for Technology, Law & Innovation. </span></em></p><p class="fine-print"><em><span>Kimberlee Weatherall receives funding from the Australian Research Council. She is a Chief Investigator with the ARC Centre of Excellence on Automated Decision-Making and Society, and a Fellow with the Gradient Institute.</span></em></p>There’s little transparency surrounding how insurance firms collect, analyse and use our personal data when they establish insurance costs.Zofia Bednarz, Lecturer in Commercial Law, University of SydneyKayleen Manwaring, Senior Research Fellow, UNSW Allens Hub for Technology, Law & Innovation and Senior Lecturer, School of Private & Commercial Law, UNSW Law & Justice, UNSW SydneyKimberlee Weatherall, Professor of Law, University of SydneyLicensed as Creative Commons – attribution, no derivatives.tag:theconversation.com,2011:article/1814132022-05-26T12:26:46Z2022-05-26T12:26:46ZGenetic mutations can be benign or cancerous – a new method to differentiate between them could lead to better treatments<figure><img src="https://images.theconversation.com/files/465132/original/file-20220524-16-83n532.jpg?ixlib=rb-1.1.0&rect=0%2C0%2C2226%2C1346&q=45&auto=format&w=496&fit=clip" /><figcaption><span class="caption">Identifying the difference between normal genetic variation and disease-causing mutations can sometimes be difficult.</span> <span class="attribution"><a class="source" href="https://www.gettyimages.com/detail/photo/gene-study-in-a-dna-chain-mutations-and-genetic-royalty-free-image/1393593249">Andrii Yalanskyi/iStock via Getty Images Plus</a></span></figcaption></figure><p>Most of the <a href="https://doi.org/10.3109/03014460.2013.807878">roughly 40 trillion cells</a> of your body have nearly identical copies of your genome – the DNA inherited from your parents, containing instructions for everything from converting food to energy to fighting off infections. Healthy cells become cancerous through <a href="https://www.cancer.net/navigating-cancer-care/cancer-basics/genetics/genetics-cancer">harmful mutations</a> in the genome. If a cell’s genome is damaged by ultraviolet light, for example, it can result in mutations that tell the cell to grow uncontrollably and form a tumor. </p>
<p>Identifying the genetic changes that cause healthy cells to become malignant can <a href="https://doi.org/10.1038/nature23306">help doctors select therapies</a> that specifically target the tumor. For example, <a href="https://doi.org/10.1097%2Fgco.0b013e3283414e87">about 25%</a> of breast cancers are <a href="https://www.cancer.org/cancer/breast-cancer/understanding-a-breast-cancer-diagnosis/breast-cancer-her2-status.html">HER2-positive</a>, meaning the cells in this type of tumor have mutations that cause them to produce more of a protein called HER2 that helps them grow. Treatments that <a href="https://doi.org/10.1016/S0140-6736(16)32417-5">specifically target</a> HER2 have dramatically increased survival rates for this type of breast cancer. </p>
<p>Scientists can now readily read cell DNA to identify mutations. The challenge is that the human genome is massive, and mutations are a normal part of evolution. The human genome is <a href="https://doi.org/10.1126%2Fscience.abj6987">long enough to fill a 1.2 million-page book</a>, and any two people can have about <a href="https://www.ashg.org/wp-content/uploads/2019/09/genetic-variation-essay.pdf">3 million genetic differences</a>. Finding one cancer-driving mutation in a tumor is like finding a needle in a stack of needles.</p>
<p>I am a <a href="https://scholar.google.com/citations?user=AXnmFTIAAAAJ&hl=en">computer scientist</a> who explores large and complex genetic data sets to answer fundamental questions about biology and disease. My research team and I <a href="https://doi.org/10.1038/s41592-022-01423-4">recently published a study</a> using DNA from thousands of healthy people to help identify disease-causing mutations by using the principle of natural selection.</p>
<figure>
<iframe width="440" height="260" src="https://www.youtube.com/embed/JWPt7AUuo-w?wmode=transparent&start=0" frameborder="0" allowfullscreen=""></iframe>
<figcaption><span class="caption">While genetic mutations are an everyday part of life, some can lead to cancer.</span></figcaption>
</figure>
<h2>Using big data to find cancerous mutations</h2>
<p>When determining what type of cancer mutation a patient has, the gold standard is to compare <a href="https://doi.org/10.1038/nrg3539">two samples from the patient</a>: one from the tumor and one from healthy tissue (typically blood). Since both samples came from the same person, most of their DNA is identical; focusing only the genetic regions that differ from each other drastically narrows the location of a possible cancer-causing mutation.</p>
<p>The problem is that healthy tissue isn’t always collected from patients, for reasons ranging from clinical costs to narrow research protocols.</p>
<p>One way to get around this is to look at massive public DNA databases. Since cancer-driving mutations are detrimental to survival, natural selection tends to eliminate them over time in successive generations. Of all the mutations in a tumor, the ones that occur less frequently in a given population are more likely to be harmful than changes that are shared by many people. By <a href="https://doi.org/10.1038/nature19057">counting how often a mutation occurs</a> in these databases, researchers can distinguish between genetic changes that are common and likely benign and those that are rare and potentially cancerous. </p>
<figure class="align-right zoomable">
<a href="https://images.theconversation.com/files/465144/original/file-20220524-21-fn1wjc.png?ixlib=rb-1.1.0&q=45&auto=format&w=1000&fit=clip"><img alt="Schematic of a cancer mutation in a cell that leads to proliferation and further genetic mutations spurring uncontrollable cell division" src="https://images.theconversation.com/files/465144/original/file-20220524-21-fn1wjc.png?ixlib=rb-1.1.0&q=45&auto=format&w=237&fit=clip" srcset="https://images.theconversation.com/files/465144/original/file-20220524-21-fn1wjc.png?ixlib=rb-1.1.0&q=45&auto=format&w=600&h=1765&fit=crop&dpr=1 600w, https://images.theconversation.com/files/465144/original/file-20220524-21-fn1wjc.png?ixlib=rb-1.1.0&q=30&auto=format&w=600&h=1765&fit=crop&dpr=2 1200w, https://images.theconversation.com/files/465144/original/file-20220524-21-fn1wjc.png?ixlib=rb-1.1.0&q=15&auto=format&w=600&h=1765&fit=crop&dpr=3 1800w, https://images.theconversation.com/files/465144/original/file-20220524-21-fn1wjc.png?ixlib=rb-1.1.0&q=45&auto=format&w=754&h=2219&fit=crop&dpr=1 754w, https://images.theconversation.com/files/465144/original/file-20220524-21-fn1wjc.png?ixlib=rb-1.1.0&q=30&auto=format&w=754&h=2219&fit=crop&dpr=2 1508w, https://images.theconversation.com/files/465144/original/file-20220524-21-fn1wjc.png?ixlib=rb-1.1.0&q=15&auto=format&w=754&h=2219&fit=crop&dpr=3 2262w" sizes="(min-width: 1466px) 754px, (max-width: 599px) 100vw, (min-width: 600px) 600px, 237px"></a>
<figcaption>
<span class="caption">One cancer-driving mutation can lead to a cascade of other mutations that lead to uncontrollable cell division.</span>
<span class="attribution"><a class="source" href="https://commons.wikimedia.org/wiki/File:Cancer_requires_multiple_mutations_from_NIHen.png">National Cancer Institute/Wikimedia Commons</a></span>
</figcaption>
</figure>
<p>Given the power of this approach, there has been a recent surge of projects to collect and share the DNA sequences from hundreds to thousands of individuals. These projects include <a href="https://www.internationalgenome.org/">the 1000 Genomes Project</a>, <a href="https://www.simonsfoundation.org/simons-genome-diversity-project/">Simons Genome Diversity Project</a>, <a href="http://gnomad.broadinstitute.org/">GnomAD</a> and <a href="https://allofus.nih.gov/">All of Us</a>. There will likely be many more in the future.</p>
<p>Estimating how likely a mutation causes disease by how frequently it appears in a genome is common for small genetic changes called <a href="https://www.garvan.org.au/research/kinghorn-centre-for-clinical-genomics/learn-about-genomics/for-gp/genetics-refresher-1/types-of-variants">single-nucleotide variants (SNVs)</a>. SNVs affect just one position in the <a href="https://medlineplus.gov/genetics/understanding/basics/dna/">3 billion neuclotide</a> human genome. It could, for example, switch one thymine T to a cytosine C.</p>
<p>Most researchers and clinical pathologists use a catalog of variants that have been detected across thousands of samples. If an SNV identified in a tumor is not listed in the catalog, we can assume that it’s rare and possibly drives cancer. This works well for SNVs because detection of these mutations is usually accurate, with few false negatives.</p>
<p>However, this process breaks down for genetic changes across longer strands of DNA called <a href="https://www.ncbi.nlm.nih.gov/dbvar/content/overview/">structural variants (SVs)</a>. SVs are more complex because they include the addition, removal, inversion or duplication of sequences. Compared to much simpler SNVs, SVs have <a href="https://doi.org/10.1002/bies.201100075">higher error rates in detection</a>. False negatives are relatively frequent, resulting in incomplete catalogs that make comparing mutations against them difficult. Finding a tumor SV that isn’t listed in a catalog could mean that it’s rare and a cancer-driving candidate, or that it was missed when the catalog was created.</p>
<h2>Focusing on verification</h2>
<p>My colleagues and I solved these problems by moving from a process focused on detection to one that focuses on verification. Detection is difficult – it requires processing complex data to determine if there is enough evidence to support the existence of a mutation. On the other hand, verification limits decision-making just to whether or not the evidence at hand supports the existence of a specific event. Instead of looking for a needle in a stack of needles, we are now simply considering whether the needle we have is the one we want.</p>
<p><a href="https://doi.org/10.1038/s41592-022-01423-4">Our method</a> leverages this strategy by searching through raw data from thousands of DNA samples for any evidence supporting specific SV. In addition to the efficiency benefits of only looking at the data flanking the target variant, if there is no such evidence, we can confidently conclude that the target variant is rare and potentially disease-causing.</p>
<p>Using our method, we scanned the SVs identified in prior cancer studies and found that thousands of SVs previously associated with cancers also appear in normal healthy samples. This indicates that these variants are more likely to be benign, inherited sequences rather than disease-causing ones.</p>
<figure>
<iframe width="440" height="260" src="https://www.youtube.com/embed/ti50nS7B5vI?wmode=transparent&start=0" frameborder="0" allowfullscreen=""></iframe>
<figcaption><span class="caption">All of Us is a research program from the National Institutes of Health with the aim of making medicine more tailored to individual needs.</span></figcaption>
</figure>
<p>Most importantly, our method performed just as well as the traditional strategy that requires both tumor and healthy samples, opening the door to reducing the cost and increasing the accessibility of high-quality cancer mutation analysis.</p>
<p>My team and I are exploring expanding our searches to include large collections of tumors from different types of cancers such as breast and lung. Determining which organ a tumor originated from is critical to prognosis and treatment because it can indicate whether the cancer has metastasized or not. Because most tumors have specific mutational signatures, recovering evidence of an SV within a specific tumor sample could help identify the patient’s tumor type and lead to faster treatment.</p><img src="https://counter.theconversation.com/content/181413/count.gif" alt="The Conversation" width="1" height="1" />
<p class="fine-print"><em><span>Ryan Layer does not work for, consult, own shares in or receive funding from any company or organization that would benefit from this article, and has disclosed no relevant affiliations beyond their academic appointment.</span></em></p>Tumors contain thousands of genetic changes, but only a few are actually cancer-causing. A quicker way to identify these driver mutations could lead to more targeted cancer treatments.Ryan Layer, Assistant Professor of Computer Science, University of Colorado BoulderLicensed as Creative Commons – attribution, no derivatives.tag:theconversation.com,2011:article/1786082022-03-23T19:06:55Z2022-03-23T19:06:55ZTwo years into the pandemic, why is Australia still short of medicines?<figure><img src="https://images.theconversation.com/files/453710/original/file-20220322-17-1qyxhns.jpg?ixlib=rb-1.1.0&rect=0%2C0%2C1000%2C666&q=45&auto=format&w=496&fit=clip" /><figcaption><span class="caption">
</span> <span class="attribution"><a class="source" href="https://www.shutterstock.com/image-photo/zaragoza-spain-december-2021-covid-test-2096748151">Alex Bascuas/Shutterstock</a></span></figcaption></figure><p>This might be a familiar scene. You pop into your local pharmacy to fill a script and you’re told your regular medicine is out of stock. When will it be in? Sorry, we don’t have a date. But I’ll ring up your GP to see if she can authorise an alternative. </p>
<p>This is a common conversation more than two years into the pandemic. So why, when our borders are open and planes are arriving from overseas with medicines on board, do we still have medicine shortages?</p>
<p>This may be surprising, but medicine shortages have been an ongoing issue in Australia. The pandemic only made it more visible. </p>
<p>For my <a href="https://vuir.vu.edu.au/view/people/Ziaee=3AMaryam=3A=3A.date.html">PhD research</a>, I looked at Australia’s pharmaceutical supply chain – the process of how medicines get from manufacturers to wholesalers and then to pharmacies.</p>
<p>I interviewed 20 supply-chain experts from 15 Australian and multinational companies. Here’s what I found, and what we could be doing better.</p>
<hr>
<p>
<em>
<strong>
Read more:
<a href="https://theconversation.com/ive-heard-covid-is-leading-to-medicine-shortages-what-can-i-do-if-my-medicine-is-out-of-stock-153628">I've heard COVID is leading to medicine shortages. What can I do if my medicine is out of stock?</a>
</strong>
</em>
</p>
<hr>
<h2>If it’s not the pandemic, what’s going on?</h2>
<p>The <a href="https://apps.tga.gov.au/Prod/msi/search?shortagetype=All">Therapeutic Goods Administration database</a> lists shortages of 263 medicines, with a critical shortage of 27 of them. Shortages of 65 more medicines are expected. The list is updated daily. </p>
<p><div data-react-class="Tweet" data-react-props="{"tweetId":"1410858531364954113"}"></div></p>
<p>However, the pandemic is not the root cause of medicine shortages. So
border openings will not solve the problem.</p>
<p>Even before the pandemic, we were regularly seeing <a href="https://pubmed.ncbi.nlm.nih.gov/27383246/">medicine shortages</a> at similar levels.</p>
<p>The pharmaceutical industry is fundamentally different from other industries. Developing medicines is an extremely lengthy process, with no guarantee of success. Some 90% of candidate drugs <a href="https://www.nature.com/articles/nrd.2016.136">don’t complete clinical trials</a>. Of those that do, <a href="https://pubmed.ncbi.nlm.nih.gov/21878973/">not all make it to market</a>. </p>
<p>Some drugs are also “<a href="https://www.sciencedirect.com/science/article/pii/S2214716016301105">personalised</a>” so they are better targeted to an individual patient’s needs. This means small quantities of tailored drugs may be needed.</p>
<p>So organisations, such as drug manufacturers, wholesalers and hospitals, mainly rely on historical data to plan the production and distribution of medicines.</p>
<figure class="align-center zoomable">
<a href="https://images.theconversation.com/files/453714/original/file-20220323-17-12n9ysk.jpg?ixlib=rb-1.1.0&q=45&auto=format&w=1000&fit=clip"><img alt="Blister packs of green capsules being made in a factory" src="https://images.theconversation.com/files/453714/original/file-20220323-17-12n9ysk.jpg?ixlib=rb-1.1.0&q=45&auto=format&w=754&fit=clip" srcset="https://images.theconversation.com/files/453714/original/file-20220323-17-12n9ysk.jpg?ixlib=rb-1.1.0&q=45&auto=format&w=600&h=341&fit=crop&dpr=1 600w, https://images.theconversation.com/files/453714/original/file-20220323-17-12n9ysk.jpg?ixlib=rb-1.1.0&q=30&auto=format&w=600&h=341&fit=crop&dpr=2 1200w, https://images.theconversation.com/files/453714/original/file-20220323-17-12n9ysk.jpg?ixlib=rb-1.1.0&q=15&auto=format&w=600&h=341&fit=crop&dpr=3 1800w, https://images.theconversation.com/files/453714/original/file-20220323-17-12n9ysk.jpg?ixlib=rb-1.1.0&q=45&auto=format&w=754&h=429&fit=crop&dpr=1 754w, https://images.theconversation.com/files/453714/original/file-20220323-17-12n9ysk.jpg?ixlib=rb-1.1.0&q=30&auto=format&w=754&h=429&fit=crop&dpr=2 1508w, https://images.theconversation.com/files/453714/original/file-20220323-17-12n9ysk.jpg?ixlib=rb-1.1.0&q=15&auto=format&w=754&h=429&fit=crop&dpr=3 2262w" sizes="(min-width: 1466px) 754px, (max-width: 599px) 100vw, (min-width: 600px) 600px, 237px"></a>
<figcaption>
<span class="caption">Until now, it’s been hard to predict which medicines need to be made and distributed to meet demand.</span>
<span class="attribution"><a class="source" href="https://www.shutterstock.com/image-photo/green-capsule-medicine-pill-production-line-1845695305">Shutterstock</a></span>
</figcaption>
</figure>
<p>However, Australia’s pharmaceutical supply chain is highly <a href="https://pubmed.ncbi.nlm.nih.gov/26244304/">fragmented</a>. There’s little coordination or data sharing among manufacturers, wholesalers and pharmacies. This leads to poor communication and incomplete or inaccurate information.</p>
<p>For instance, manufacturers may have little or no access to pharmacy data and stock levels. So, they cannot adequately plan for medicine production, which can take from several months to a year.</p>
<p>Australia also accounts for only <a href="https://www.ncbi.nlm.nih.gov/pmc/articles/PMC6202294/">2% of the world’s drug market</a>, a small one for
multinational pharmaceutical manufacturers. So their domestic suppliers generally keep a low stock due to short expiry dates and profit margins. </p>
<p>The slightest disruption, such as disease outbreaks or natural disasters, can easily spike demands and cause a shortage. Pandemic-related supply chain disruptions only make the existing challenges worse.</p>
<hr>
<p>
<em>
<strong>
Read more:
<a href="https://theconversation.com/supermarket-shortages-are-different-this-time-how-to-respond-and-avoid-panic-174529">Supermarket shortages are different this time: how to respond and avoid panic</a>
</strong>
</em>
</p>
<hr>
<h2>What can we do about it?</h2>
<p>If a medicine is in short supply, there may be an alternative option a doctor can prescribe. But substituting medicines <a href="https://journals.plos.org/plosone/article?id=10.1371/journal.pone.0215837">can lead to</a> side effects, longer recovery times, longer stays in hospitals, and increased health-care costs. </p>
<p>Some pharmacies and wholesalers overstock their warehouses if they anticipate a shortage. But that is costly and medicines might expire before they are used. These are only short-term solutions.</p>
<p>So we need a system-wide and nationally coordinated approach among supply chain partners and the government to reduce the risk of medicine shortages.</p>
<figure class="align-center ">
<img alt="" src="https://images.theconversation.com/files/453557/original/file-20220322-24-18n7zya.jpg?ixlib=rb-1.1.0&q=45&auto=format&w=754&fit=clip" srcset="https://images.theconversation.com/files/453557/original/file-20220322-24-18n7zya.jpg?ixlib=rb-1.1.0&q=45&auto=format&w=600&h=338&fit=crop&dpr=1 600w, https://images.theconversation.com/files/453557/original/file-20220322-24-18n7zya.jpg?ixlib=rb-1.1.0&q=30&auto=format&w=600&h=338&fit=crop&dpr=2 1200w, https://images.theconversation.com/files/453557/original/file-20220322-24-18n7zya.jpg?ixlib=rb-1.1.0&q=15&auto=format&w=600&h=338&fit=crop&dpr=3 1800w, https://images.theconversation.com/files/453557/original/file-20220322-24-18n7zya.jpg?ixlib=rb-1.1.0&q=45&auto=format&w=754&h=424&fit=crop&dpr=1 754w, https://images.theconversation.com/files/453557/original/file-20220322-24-18n7zya.jpg?ixlib=rb-1.1.0&q=30&auto=format&w=754&h=424&fit=crop&dpr=2 1508w, https://images.theconversation.com/files/453557/original/file-20220322-24-18n7zya.jpg?ixlib=rb-1.1.0&q=15&auto=format&w=754&h=424&fit=crop&dpr=3 2262w" sizes="(min-width: 1466px) 754px, (max-width: 599px) 100vw, (min-width: 600px) 600px, 237px">
<figcaption>
<span class="caption">Looking at vast amounts of data, from many sources, in real time is the key.</span>
<span class="attribution"><span class="source">Shutterstock</span></span>
</figcaption>
</figure>
<p>We could do this using artificial intelligence technologies such as “<a href="https://www.sciencedirect.com/science/article/abs/pii/S1366554516303799?via%3Dihub">big data analytics</a>” and sharing data across the pharmaceutical supply chain. </p>
<p>Big data analytics can store and analyse a large array of data in different formats, from different sources, in real time. This would create an integrated database for all pharmaceutical supply-chain partners to have access to. This would allow all partners to monitor and predict demand in real time.</p>
<p>For example, a pharmacist would be able to access a centralised database on their computer and check the current stock level and availability of a medicine in other pharmacies, or even manufacturers and distributors. This could even help predict medicine shortages way before they occur. </p>
<p>For this to work, Australian pharmaceutical organisations need both robust IT and a skilled workforce that knows how to analyse and use the data. While this might be practical and affordable for pharmaceutical companies, this might not be the case for hospital or community pharmacies. </p>
<p>So governments would <a href="https://www.pc.gov.au/inquiries/completed/data-access/report/data-access.pdf">need to support</a> pharmacies and other smaller players – technically, financially, and with appropriate policies and regulations – to make sure they could access and use the data.</p>
<h2>We need to plan for the next crisis</h2>
<p>The current pandemic may be adding additional stresses to an already stretched supply chain. But future pandemics and natural disasters, such as floods and bushfires, will also worsen medicine shortages.</p>
<p>So we need to start planning now to create a resilient pharmaceutical supply chain that predicts medicine shortages and responds quickly.</p><img src="https://counter.theconversation.com/content/178608/count.gif" alt="The Conversation" width="1" height="1" />
<p class="fine-print"><em><span>Maryam Ziaee does not work for, consult, own shares in or receive funding from any company or organisation that would benefit from this article, and has disclosed no relevant affiliations beyond their academic appointment.</span></em></p>It’s easy to blame COVID. But Australia has suffered medicine shortages for years. The pandemic has only highlighted the problem. Here’s what we could do to better avoid shortages in the first place.Maryam Ziaee, Lecturer and Researcher, Supply Chains and Operations, Victoria UniversityLicensed as Creative Commons – attribution, no derivatives.tag:theconversation.com,2011:article/1763472022-03-16T16:34:27Z2022-03-16T16:34:27ZHow AI helped deliver cash aid to many of the poorest people in Togo<figure><img src="https://images.theconversation.com/files/451998/original/file-20220314-131648-9brlb3.jpg?ixlib=rb-1.1.0&rect=198%2C68%2C4825%2C3015&q=45&auto=format&w=496&fit=clip" /><figcaption><span class="caption">Mobile devices are becoming ubiquitous in Africa.</span> <span class="attribution"><a class="source" href="https://www.gettyimages.com/detail/news-photo/mobile-phone-lome-togo-news-photo/170481943">Godong/Universal Images Group via Getty Images</a></span></figcaption></figure><p><em>The <a href="https://theconversation.com/us/topics/research-brief-83231">Research Brief</a> is a short take about interesting academic work.</em></p>
<h2>The big idea</h2>
<p>Governments and humanitarian groups can use machine learning algorithms and mobile phone data to <a href="https://doi.org/10.1038/s41586-022-04484-9">get aid to those who need it most</a> during a humanitarian crisis, we found in new research.</p>
<figure class="align-left zoomable">
<a href="https://images.theconversation.com/files/446801/original/file-20220216-24-1v07seb.jpg?ixlib=rb-1.1.0&q=45&auto=format&w=1000&fit=clip"><img alt="Map of West Africa, highlighting Togo" src="https://images.theconversation.com/files/446801/original/file-20220216-24-1v07seb.jpg?ixlib=rb-1.1.0&q=45&auto=format&w=237&fit=clip" srcset="https://images.theconversation.com/files/446801/original/file-20220216-24-1v07seb.jpg?ixlib=rb-1.1.0&q=45&auto=format&w=600&h=556&fit=crop&dpr=1 600w, https://images.theconversation.com/files/446801/original/file-20220216-24-1v07seb.jpg?ixlib=rb-1.1.0&q=30&auto=format&w=600&h=556&fit=crop&dpr=2 1200w, https://images.theconversation.com/files/446801/original/file-20220216-24-1v07seb.jpg?ixlib=rb-1.1.0&q=15&auto=format&w=600&h=556&fit=crop&dpr=3 1800w, https://images.theconversation.com/files/446801/original/file-20220216-24-1v07seb.jpg?ixlib=rb-1.1.0&q=45&auto=format&w=754&h=699&fit=crop&dpr=1 754w, https://images.theconversation.com/files/446801/original/file-20220216-24-1v07seb.jpg?ixlib=rb-1.1.0&q=30&auto=format&w=754&h=699&fit=crop&dpr=2 1508w, https://images.theconversation.com/files/446801/original/file-20220216-24-1v07seb.jpg?ixlib=rb-1.1.0&q=15&auto=format&w=754&h=699&fit=crop&dpr=3 2262w" sizes="(min-width: 1466px) 754px, (max-width: 599px) 100vw, (min-width: 600px) 600px, 237px"></a>
<figcaption>
<span class="caption">Togo is a small West African nation.</span>
<span class="attribution"><a class="source" href="https://www.gettyimages.com/detail/news-photo/world-data-locator-map-togo-news-photo/641462678">Encyclopaedia Britannica/Getty Images</a></span>
</figcaption>
</figure>
<p>The <a href="https://doi.org/10.1126/science.aac4420">simple idea</a> behind this approach, as we explained in the journal <a href="https://doi.org/10.1038/s41586-022-04484-9">Nature on March 16, 2022</a>, is that wealthy people use phones differently from poor people. Their phone calls and text messages follow different patterns, and they use different data plans, for example. Machine learning algorithms – which are fancy tools for pattern recognition – can be trained to recognize those differences and infer whether a given mobile subscriber is wealthy or poor.</p>
<p>As the COVID-19 pandemic spread in early 2020, <a href="https://www.ischool.berkeley.edu/">our</a> <a href="https://www.poverty-action.org/">research</a> <a href="https://cega.berkeley.edu/">team</a> helped Togo’s <a href="https://numerique.gouv.tg/">Ministry of Digital Economy</a> and <a href="https://www.givedirectly.org/">GiveDirectly</a>, a nonprofit that sends cash to people living in poverty, turn this insight into a new type of aid program. </p>
<p>First, we collected recent, reliable and representative data. Working on the ground with partners in Togo, we conducted 15,000 phone surveys to collect information on the living conditions of each household. After matching the survey responses to data from the mobile phone companies, we trained the machine learning algorithms to recognize the patterns of phone use that were characteristics of people living on less than $1.25 per day.</p>
<p>The next challenge was figuring out whether a system based on machine learning and phone data would be effective at getting money to the poorest people in the country. <a href="https://doi.org/10.1038/s41586-022-04484-9">Our evaluation</a> indicated that this new approach worked better than other options Togo’s government was considering.</p>
<p>For instance, focusing entirely on the poorest cantons – which are analagous to U.S. counties – would have delivered benefits to only 33% of the people living on less than US$1.25 a day. By contrast, the machine learning approach targeted 47% of that population.</p>
<p>We then partnered with Togo’s government, GiveDirectly and community leaders to design and pilot a cash transfer program based on this technology. In November 2020, the first beneficiaries were <a href="https://www.bbc.com/news/av/stories-56580833">enrolled and paid</a>. To date, the program has provided nearly $10 million to roughly 137,000 of the country’s poorest citizens. </p>
<h2>Why it matters</h2>
<p>Our work shows that data collected by mobile phone companies – when analyzed with machine learning technology – can help <a href="https://www.npr.org/sections/goatsandsoda/2021/02/15/966848542/the-pandemic-pushed-this-farmer-into-deep-poverty-then-something-amazing-happene">direct aid</a> to those with the greatest need.</p>
<p>Even before the pandemic, over half of the <a href="https://data.worldbank.org/indicator/NY.GDP.PCAP.CD?locations=TG">West African nation’s</a> 8.6 million people lived below the international poverty line. As COVID-19 slowed economic activity further, our surveys indicated that 54% of all Togolese were forced to miss meals each week.</p>
<p>The situation in Togo was not unique. The downturn resulting from the COVID-19 pandemic <a href="https://blogs.worldbank.org/opendata/updated-estimates-impact-covid-19-global-poverty-turning-corner-pandemic-2021">pushed millions of people into extreme poverty</a>. In response, governments and charities launched several thousand new aid programs, providing benefits to <a href="https://openknowledge.worldbank.org/handle/10986/33635">over 1.5 billion people and families</a> around the world. </p>
<p>But in the middle of a humanitarian crisis, governments struggle to figure out who needs help most urgently. Under ideal circumstances, those decisions would be based on comprehensive household surveys. But there was no way to gather this information in the middle of a pandemic.</p>
<p>Our work helps demonstrate how new sources of big data – such as information gleaned from satellites and mobile phone networks – can make it possible to target aid amid crisis conditions when more traditional sources of data are unavailable. </p>
<h2>What’s next</h2>
<p>We’re conducting follow-up research to assess how cash transfers affected recipients. <a href="https://doi.org/10.1093/qje/qjw025">Previous</a> <a href="https://doi.org/10.3386/w26600">findings</a> indicate that cash transfers can help increase food security and improve psychological well-being in normal times. We are assessing whether that aid has similar results during a crisis.</p>
<p>It’s also essential to find ways to enroll and pay people without phones. In Togo, roughly 85% of households had at least one phone, and <a href="https://doi.org/10.48550/arXiv.2111.00175">phones are frequently shared</a> within families and communities. However, it is not clear how many people who needed humanitarian assistance in Togo didn’t get it because of their lack of access to a mobile device.</p>
<p>In the future, systems that combine new methods that leverage machine learning and big data with traditional approaches based on surveys are bound to improve the targeting of humanitarian aid.</p>
<p>[<em>Over 150,000 readers rely on The Conversation’s newsletters to understand the world.</em> <a href="https://memberservices.theconversation.com/newsletters/?source=inline-150ksignup">Sign up today</a>.]</p><img src="https://counter.theconversation.com/content/176347/count.gif" alt="The Conversation" width="1" height="1" />
<p class="fine-print"><em><span>Emily Aiken collaborated closely with the teams at GiveDirectly and the government of Togo described in the article. She consulted for GiveDirectly from June to August 2021. </span></em></p><p class="fine-print"><em><span>Joshua Blumenstock receives funding from Google.org, data.org, the Center for Effective Global Action, the Jameel Poverty Action Lab, and the NSF under award IIS – 1942702.
</span></em></p>To date, the program has provided nearly $10 million to roughly 137,000 of the country’s poorest citizens.Emily Aiken, Doctoral Student of Information, University of California, BerkeleyJoshua Blumenstock, Associate Professor of Information; Co-Director of the Center for Effective Global Action, University of California, BerkeleyLicensed as Creative Commons – attribution, no derivatives.tag:theconversation.com,2011:article/1768912022-03-07T19:06:45Z2022-03-07T19:06:45ZHow we communicate, what we value – even who we are: 8 surprising things data science has revealed about us over the past decade<figure><img src="https://images.theconversation.com/files/450290/original/file-20220307-84591-em3wfo.jpeg?ixlib=rb-1.1.0&rect=249%2C62%2C6674%2C4556&q=45&auto=format&w=496&fit=clip" /><figcaption><span class="caption">
</span> <span class="attribution"><span class="source">Shutterstock</span></span></figcaption></figure><p>Big data analysis has long supported <a href="https://www.newscientist.com/article/mg24432613-200-new-scientist-ranks-the-top-10-discoveries-of-the-decade/">major feats</a> in physics and astronomy. But more recently we’ve seen it underpin breakthroughs in the social sciences and humanities.</p>
<p>Since the landmark paper <a href="https://www.science.org/doi/10.1126/science.1167742">Computational Social Science</a> was published in 2009, a new generation of data analytics tools has given researchers insight into fundamental questions about how we communicate, who we are and what we value. </p>
<p>For instance, by analysing the relative frequency of certain words in historical texts, researchers can identify important changes in our use of language over time.</p>
<p>In some cases these shifts will be obvious, such as the use of archaic words being replaced by more contemporary words. But in other cases, they may reflect more subtle but widespread social and cultural changes. Below are some of the most influential data-centric discoveries from the past 10 years.</p>
<h2>How we communicate</h2>
<p>Over the past decade, a growing number of global open data sources have helped researchers reveal patterns in what we read, write and pay attention to. Google Books, <a href="https://www.worldcat.org/">Worldcat</a> and <a href="https://www.gutenberg.org/">Project Gutenberg</a> are just some examples.</p>
<p>The release of the Google Books <a href="https://books.google.com/ngrams">n-gram viewer</a> in the early 2010s was a game changer on this front. Using the entire Google Books database, this tool shows you the relative frequency of a specific term or phrase as it has been used over hundreds of years. <a href="https://www.scientificamerican.com/article/google-books-culture/">Researchers</a> have used this data to explore the systematic suppression of the mention of Jewish painters, such as Marc Chagall, in German books during World War II.</p>
<p>Data analysis can also reveal patterns in the expression of human emotions over time. CSIRO’s <a href="http://wefeel.csiro.au/">We Feel</a> tracks emotions in communities around the world. It does this by analysing the language people are using on social media in real time and mapping it out. </p>
<p>The tool can be used to determine the general mood over time (hour by hour, day by day) within particular cities and countries. Patterns in these data can then be explored in association with other information, such as weather, holidays and economic fluctuations. </p>
<p>Some research findings even claim to represent fundamental changes in humans’ social values, community sentiment and how we think (for example, the rise and fall of words associated with rationality such as “method”, “analysis” and “determine”).</p>
<p>Here are some key findings in this space:
</p><ul>
<li> <strong>Cultural turnover is accelerating</strong> <p></p>
<p>A Harvard University-led <a href="https://www.science.org/doi/10.1126/science.1199644">analysis</a> of more than a century of data from millions of books provides evidence that society’s attention span for historical events is declining, as appetite for new material grows. </p>
<p>In other words, we are forgetting the past faster. You can see this in the graph below, which tracks how often three specific years are mentioned across a vast range of literature through time. As time passes, the “half-life” of each year (the point at which it receives just half the attention it had at its peak) comes quicker.</p>
<figure class="align-center ">
<img alt="Counts of mentions of the years 1883, 1910 and 1950 in all books for the past 200 years." src="https://images.theconversation.com/files/449139/original/file-20220301-15-1kjxrze.jpg?ixlib=rb-1.1.0&q=45&auto=format&w=754&fit=clip" srcset="https://images.theconversation.com/files/449139/original/file-20220301-15-1kjxrze.jpg?ixlib=rb-1.1.0&q=45&auto=format&w=600&h=395&fit=crop&dpr=1 600w, https://images.theconversation.com/files/449139/original/file-20220301-15-1kjxrze.jpg?ixlib=rb-1.1.0&q=30&auto=format&w=600&h=395&fit=crop&dpr=2 1200w, https://images.theconversation.com/files/449139/original/file-20220301-15-1kjxrze.jpg?ixlib=rb-1.1.0&q=15&auto=format&w=600&h=395&fit=crop&dpr=3 1800w, https://images.theconversation.com/files/449139/original/file-20220301-15-1kjxrze.jpg?ixlib=rb-1.1.0&q=45&auto=format&w=754&h=496&fit=crop&dpr=1 754w, https://images.theconversation.com/files/449139/original/file-20220301-15-1kjxrze.jpg?ixlib=rb-1.1.0&q=30&auto=format&w=754&h=496&fit=crop&dpr=2 1508w, https://images.theconversation.com/files/449139/original/file-20220301-15-1kjxrze.jpg?ixlib=rb-1.1.0&q=15&auto=format&w=754&h=496&fit=crop&dpr=3 2262w" sizes="(min-width: 1466px) 754px, (max-width: 599px) 100vw, (min-width: 600px) 600px, 237px">
<figcaption>
<span class="caption">Our collective attention for historical events has shrunk over the past century.</span>
<span class="attribution"><a class="source" href="https://www.science.org/doi/10.1126/science.1199644">Michel et al., Science 2010</a></span>
</figcaption>
</figure>
<p></p></li><li> <strong>Human language diversity and biodiversity are correlated</strong><p></p>
<p>By mapping linguistic diversity and the diversity of animal species, researchers have <a href="https://doi.org/10.1038/s41598-020-76658-2">shown</a> these two worlds are correlated geographically – both increasing with temperature and proximity to the equator. So the closer to the equator you get, the more variation there is in spoken language and the greater the variety of species there is. </p>
<p>The authors propose this is due to heat near the equator producing greater productivity and variety in plant life, which in turn provides more complex and interactive environments for both animals and humans alike – feeding into a cycle whereby “diversity begets more diversity”.</p>
<figure class="align-center zoomable">
<a href="https://images.theconversation.com/files/449991/original/file-20220304-36214-pyfb2r.jpg?ixlib=rb-1.1.0&q=45&auto=format&w=1000&fit=clip"><img alt="Three figures showing diversity distributions of language and animals and their relation to geography." src="https://images.theconversation.com/files/449991/original/file-20220304-36214-pyfb2r.jpg?ixlib=rb-1.1.0&q=45&auto=format&w=754&fit=clip" srcset="https://images.theconversation.com/files/449991/original/file-20220304-36214-pyfb2r.jpg?ixlib=rb-1.1.0&q=45&auto=format&w=600&h=503&fit=crop&dpr=1 600w, https://images.theconversation.com/files/449991/original/file-20220304-36214-pyfb2r.jpg?ixlib=rb-1.1.0&q=30&auto=format&w=600&h=503&fit=crop&dpr=2 1200w, https://images.theconversation.com/files/449991/original/file-20220304-36214-pyfb2r.jpg?ixlib=rb-1.1.0&q=15&auto=format&w=600&h=503&fit=crop&dpr=3 1800w, https://images.theconversation.com/files/449991/original/file-20220304-36214-pyfb2r.jpg?ixlib=rb-1.1.0&q=45&auto=format&w=754&h=633&fit=crop&dpr=1 754w, https://images.theconversation.com/files/449991/original/file-20220304-36214-pyfb2r.jpg?ixlib=rb-1.1.0&q=30&auto=format&w=754&h=633&fit=crop&dpr=2 1508w, https://images.theconversation.com/files/449991/original/file-20220304-36214-pyfb2r.jpg?ixlib=rb-1.1.0&q=15&auto=format&w=754&h=633&fit=crop&dpr=3 2262w" sizes="(min-width: 1466px) 754px, (max-width: 599px) 100vw, (min-width: 600px) 600px, 237px"></a>
<figcaption>
<span class="caption">Researchers have shown both linguistic diversity and species diversity increase exponentially with temperature and proximity to the equator.</span>
<span class="attribution"><a class="source" href="https://www.nature.com/articles/s41598-020-76658-2">Hamilton, Walker & Kempes, Scientific Reports 2020</a></span>
</figcaption>
</figure>
<p></p></li><li><strong>There have been society-wide shifts in language use over the past century</strong><p></p>
<p>In an article <a href="https://www.pnas.org/doi/10.1073/pnas.2107848118">published</a> in December researchers used machine learning to show long-term, consistent changes in our use of language. Specifically, they reveal an inflection point in the 1980s where there is a shift towards more egocentric, emotional and supposedly less rational language.</p>
<p>The authors suggest (although not <a href="https://doi.org/10.1073/pnas.2121300119">without contest</a>) this could signal the beginning of a “post-truth era”.
</p></li></ul><p></p>
<h2>Who we are</h2>
<p>In the field of psychology, the same data analytics tools have shown that people’s personalities can be measured using the “Big 5” traits, which largely become <a href="https://doi.org/10.1016/j.econlet.2011.11.015">stable in adulthood</a>.</p>
<p>This was possible thanks to extensive data sets such as HILDA in Australia, the German Socio-Economic Panel in Germany and the British Household Panel Survey in the UK. </p>
<p>Robust studies have also demonstrated that personality traits can be reliably and accurately predicted from a variety of data sources including <a href="https://doi.org/10.1027/1614-0001/a000362">voice recordings</a>, <a href="https://doi.org/10.1073/pnas.1920484117">mobile phone usage patterns</a> and even <a href="https://www.nature.com/articles/s41598-020-79310-1">portrait photographs</a>. </p>
<p>In turn, there have been some remarkable associations found at scale between personality and:
</p><ul>
<li> <strong>Elevation</strong><p></p>
<p>A study published in 2020, and based on more than three million people’s data, <a href="https://www.nature.com/articles/s41562-020-0930-x">shows</a> mountain-dwelling people tend to have different personality traits than those who live at sea level. They are generally more open to new experiences and more emotionally stable.</p>
<p></p></li><li> <strong>Location</strong> <p></p>
<p>Another earlier study shows people who live in the United States can be divided into <a href="https://www.apa.org/pubs/journals/releases/psp-a0034434.pdf">three clear and measurable clusters</a> of personality types, linked with associated geographic footprints. New Yorkers and Texans (who are in the same cluster) are more likely to be temperamental and uninhibited.</p>
<p></p></li><li> <strong>Occupation</strong><p></p>
<p>In our own research published with colleagues in 2019, we analysed the personality features of people in more than 1,000 different occupations. <a href="https://www.pnas.org/doi/10.1073/pnas.1917942116">We found</a> people in the same role share similar traits. Scientists are more open to new ideas yet <a href="https://www.natureindex.com/news-blog/scientists-are-curious-and-idealistic-but-not-very-agreeable-compared-to-other-professions">ready to argue</a>, whereas tennis professionals tend to be friendly and outgoing. </p>
<p>The research used machine learning to infer the personality features of more than 100,000 people, based on language used on social media.</p>
<p></p></li></ul> <p></p>
<hr>
<p>
<em>
<strong>
Read more:
<a href="https://theconversation.com/robot-career-advisor-ai-may-soon-be-able-to-analyse-your-tweets-to-match-you-to-a-job-128777">Robot career advisor: AI may soon be able to analyse your tweets to match you to a job</a>
</strong>
</em>
</p>
<hr>
<h2>What we value</h2>
<p>In economics, we’re seeing major research frontiers being opened up thanks to data analysis, including in:
</p><ul>
<li> <strong>Network science</strong> <p></p>
<p>When it comes to success, we’ve learnt that performance matters most when it can be measured (like in sport). But in other fields where it can’t be measured easily (like in the art world), networks <a href="https://doi.org/10.1140/epjds/s13688-020-00227-w">matter</a> <a href="https://www.science.org/doi/10.1126/science.aau7224">most</a>. </p>
<p></p></li><li> <strong>Behavioural economics</strong> <p></p>
<p>We can now see how we behave as individuals <em>en masse</em>, unveiling valuable clues for effective policy interventions around employment, taxation and education. For instance, one <a href="https://journals.plos.org/plosone/article?id=10.1371/journal.pone.0128692">large-scale study</a> revealed those quickest to re-enter the workforce displayed certain key behaviours. These included being an early riser and being geographically mobile (perhaps meaning they’re more willing to travel further, or relocate, for work).
</p></li></ul> <p></p>
<h2>Post-theory science?</h2>
<p>Some have argued data science poses a fundamental challenge to the traditional sciences, with the emergence of “<a href="https://www.theguardian.com/technology/2022/jan/09/are-we-witnessing-the-dawn-of-post-theory-science">post-theory science</a>”. This is the concept that machines are better at understanding the relationship between data and reality than the traditional scientific method of <em>hypothesise, predict and test</em>. </p>
<p>However, reports of the <a href="https://www.wired.com/2008/06/pb-theory/">death of theory</a> are perhaps greatly exaggerated. Data are not perfect. And data science based on incomplete or biased data has the potential to miss, or mask, important patterns in human activity. This can only be addressed by critical thinking and theory. </p>
<hr>
<p>
<em>
<strong>
Read more:
<a href="https://theconversation.com/nobel-economics-prize-winners-showed-economists-how-to-turn-the-real-world-into-their-laboratory-169697">Nobel economics prize winners showed economists how to turn the real world into their laboratory</a>
</strong>
</em>
</p>
<hr>
<img src="https://counter.theconversation.com/content/176891/count.gif" alt="The Conversation" width="1" height="1" />
<p class="fine-print"><em><span>The authors do not work for, consult, own shares in or receive funding from any company or organisation that would benefit from this article, and have disclosed no relevant affiliations beyond their academic appointment.</span></em></p>Big data analysis has unveiled startling links between seemingly unrelated things, such as how a person’s physical elevation above sea level might influence their personality.Paul X. McCarthy, Adjunct Professor, UNSW SydneyColin Griffith, Strategy & Business Development, Data61, CSIROLicensed as Creative Commons – attribution, no derivatives.tag:theconversation.com,2011:article/1765672022-02-21T23:55:12Z2022-02-21T23:55:12ZA new exhibition explores invisible data, from facial algorithms to satellite tracking as a return to Country<figure><img src="https://images.theconversation.com/files/447674/original/file-20220221-27005-1czmmks.jpg?ixlib=rb-1.1.0&rect=42%2C8%2C2785%2C4226&q=45&auto=format&w=496&fit=clip" /><figcaption><span class="caption">
</span> <span class="attribution"><span class="source">Topbunk</span></span></figcaption></figure><p><em>Review: Invisibility, MOD museum, Adelaide</em></p>
<p>Disinformation, algorithms, big data, care work, climate change, cultural knowledge: they can all be invisible.</p>
<p>In her New York Times bestseller, <a href="https://www.goodreads.com/book/show/28186015-weapons-of-math-destruction">Weapons of Math Destruction</a> (2016), subtitled “how big data increases inequality and threatens democracy”, mathematician and data scientist Cathy O’Neil unpacks the elusive algorithms of our everyday lives and how accurate, fair or biased they might be. </p>
<p>Algorithms hide behind the assumed objectivity of maths, but they very much contain the biases, subjective decisions and cultural frameworks of those who design them. With scant detail on how these algorithms are created, O’Neil describes them as “inscrutable black boxes”. </p>
<p>Opaqueness is intentional. </p>
<p>In one of the upstairs galleries at the spacious MOD, we are greeted in large text as we enter: “what do algorithms think about you?” </p>
<p>Can an algorithm think?, we ask. And, if so, what informs the decisions it makes about us? </p>
<p><a href="https://biometricmirror.com/">Biometric Mirror</a> was created by science fiction artist Lucy McRae and computer scientists Niels Wouters and Frank Vetere. They created an algorithm to judge our personalities by asking 33,000 people to look at photographs of faces and come up with possible personality traits. </p>
<figure class="align-center zoomable">
<a href="https://images.theconversation.com/files/447677/original/file-20220221-25-h3frex.jpg?ixlib=rb-1.1.0&q=45&auto=format&w=1000&fit=clip"><img alt="A man looks at his reflection." src="https://images.theconversation.com/files/447677/original/file-20220221-25-h3frex.jpg?ixlib=rb-1.1.0&q=45&auto=format&w=754&fit=clip" srcset="https://images.theconversation.com/files/447677/original/file-20220221-25-h3frex.jpg?ixlib=rb-1.1.0&q=45&auto=format&w=600&h=401&fit=crop&dpr=1 600w, https://images.theconversation.com/files/447677/original/file-20220221-25-h3frex.jpg?ixlib=rb-1.1.0&q=30&auto=format&w=600&h=401&fit=crop&dpr=2 1200w, https://images.theconversation.com/files/447677/original/file-20220221-25-h3frex.jpg?ixlib=rb-1.1.0&q=15&auto=format&w=600&h=401&fit=crop&dpr=3 1800w, https://images.theconversation.com/files/447677/original/file-20220221-25-h3frex.jpg?ixlib=rb-1.1.0&q=45&auto=format&w=754&h=504&fit=crop&dpr=1 754w, https://images.theconversation.com/files/447677/original/file-20220221-25-h3frex.jpg?ixlib=rb-1.1.0&q=30&auto=format&w=754&h=504&fit=crop&dpr=2 1508w, https://images.theconversation.com/files/447677/original/file-20220221-25-h3frex.jpg?ixlib=rb-1.1.0&q=15&auto=format&w=754&h=504&fit=crop&dpr=3 2262w" sizes="(min-width: 1466px) 754px, (max-width: 599px) 100vw, (min-width: 600px) 600px, 237px"></a>
<figcaption>
<span class="caption">Can an algorithm tell you who you really are?</span>
<span class="attribution"><span class="source">Topbunk</span></span>
</figcaption>
</figure>
<p>We don’t see who the photos are of or who is doing the evaluating – and therefore we don’t know what biases might be reproduced. </p>
<p>You are invited to gaze into a mirror which scans your face. From this scan, the algorithm creates a profile of your age, gender and personality, which appears as statistical data overlaid on your reflection. </p>
<p>When I look into the mirror, I am told I am neither trustworthy nor emotionally stable. The algorithm underneath under-guesses my age by a few years, and I score highly for intelligence and uncertainty – an unhelpful combination. </p>
<p>Despite my doubts about the algorithm, I notice myself focusing on the more favourable data.</p>
<p>In this context, the data is benign. But facial recognition technology has been used to survey and monitor activists and has been responsible for <a href="https://www.theguardian.com/uk-news/2018/may/05/welsh-police-wrongly-identify-thousands-as-potential-criminals">thousands of inaccurate identifications</a> by police in the UK.</p>
<hr>
<p>
<em>
<strong>
Read more:
<a href="https://theconversation.com/algorithms-can-decide-your-marks-your-work-prospects-and-your-financial-security-how-do-you-know-theyre-fair-171590">Algorithms can decide your marks, your work prospects and your financial security. How do you know they're fair?</a>
</strong>
</em>
</p>
<hr>
<h2>Using data to illuminate cultural knowledge</h2>
<p>In one of the more impressive works in the exhibition, contemporary data visualisation is used to illustrate Aboriginal forms of knowing and the intrinsic relationship between spatial awareness, Country and kinship. </p>
<p><a href="https://trackerdataproject.com/">Ngapulara Ngarngarnyi Wirra (Our Family Tree)</a> is a collaboration between Angie Abdilla from design agency Old Ways, New, Adnyamathanha and Narungga footballer Adam Goodes and contemporary artist Baden Pailthorpe.</p>
<p>In every AFL game Goodes played, his on-field movements was recorded via satellites, which connected with a tracking device in the back of his jumper. 20 million data points were then fused with data scans of a Red River Gum, or <em>Wirra</em>, to form an impressive data visualisation projected onto two large screens in a darkened gallery. </p>
<figure class="align-center zoomable">
<a href="https://images.theconversation.com/files/447676/original/file-20220221-19-spdizl.jpg?ixlib=rb-1.1.0&q=45&auto=format&w=1000&fit=clip"><img alt="A large screen with swirling earthy colours." src="https://images.theconversation.com/files/447676/original/file-20220221-19-spdizl.jpg?ixlib=rb-1.1.0&q=45&auto=format&w=754&fit=clip" srcset="https://images.theconversation.com/files/447676/original/file-20220221-19-spdizl.jpg?ixlib=rb-1.1.0&q=45&auto=format&w=600&h=401&fit=crop&dpr=1 600w, https://images.theconversation.com/files/447676/original/file-20220221-19-spdizl.jpg?ixlib=rb-1.1.0&q=30&auto=format&w=600&h=401&fit=crop&dpr=2 1200w, https://images.theconversation.com/files/447676/original/file-20220221-19-spdizl.jpg?ixlib=rb-1.1.0&q=15&auto=format&w=600&h=401&fit=crop&dpr=3 1800w, https://images.theconversation.com/files/447676/original/file-20220221-19-spdizl.jpg?ixlib=rb-1.1.0&q=45&auto=format&w=754&h=504&fit=crop&dpr=1 754w, https://images.theconversation.com/files/447676/original/file-20220221-19-spdizl.jpg?ixlib=rb-1.1.0&q=30&auto=format&w=754&h=504&fit=crop&dpr=2 1508w, https://images.theconversation.com/files/447676/original/file-20220221-19-spdizl.jpg?ixlib=rb-1.1.0&q=15&auto=format&w=754&h=504&fit=crop&dpr=3 2262w" sizes="(min-width: 1466px) 754px, (max-width: 599px) 100vw, (min-width: 600px) 600px, 237px"></a>
<figcaption>
<span class="caption">In Ngapulara Ngarngarnyi Wirra (Our Family Tree), data from Adam Goodes’ football games is returned to Country.</span>
<span class="attribution"><span class="source">Topbunk</span></span>
</figcaption>
</figure>
<p>Here, Goodes’s data is returned to Country to form part of the roots of the tree as well as the swirling North and South Winds of his ancestors. The data is also translated into sound and amplified, inviting us to listen to what would otherwise be inaudible.</p>
<p>In a small room between the screens – or within the tree – drone footage of the Adnyamathanha Country (Flinders Ranges) plays against the retelling of the creation story in Adnyamathanha language. </p>
<p>What results is the synthesis of traditional Aboriginal knowledge with cutting edge technology, revealing different ways of sensing space and time. </p>
<hr>
<p>
<em>
<strong>
Read more:
<a href="https://theconversation.com/the-land-we-play-on-equality-doesnt-mean-justice-62101">The land we play on: equality doesn’t mean justice</a>
</strong>
</em>
</p>
<hr>
<h2>The power of the invisible</h2>
<p>While it’s easy to focus on how technology is used and exposed in the works in Invisibility, down the corridors and hanging from the ceiling in MOD are a few other exhibits that flesh out the concept of invisibility. </p>
<figure class="align-center zoomable">
<a href="https://images.theconversation.com/files/447675/original/file-20220221-22-1ecl4d9.jpg?ixlib=rb-1.1.0&q=45&auto=format&w=1000&fit=clip"><img alt="Black and white portraits" src="https://images.theconversation.com/files/447675/original/file-20220221-22-1ecl4d9.jpg?ixlib=rb-1.1.0&q=45&auto=format&w=754&fit=clip" srcset="https://images.theconversation.com/files/447675/original/file-20220221-22-1ecl4d9.jpg?ixlib=rb-1.1.0&q=45&auto=format&w=600&h=898&fit=crop&dpr=1 600w, https://images.theconversation.com/files/447675/original/file-20220221-22-1ecl4d9.jpg?ixlib=rb-1.1.0&q=30&auto=format&w=600&h=898&fit=crop&dpr=2 1200w, https://images.theconversation.com/files/447675/original/file-20220221-22-1ecl4d9.jpg?ixlib=rb-1.1.0&q=15&auto=format&w=600&h=898&fit=crop&dpr=3 1800w, https://images.theconversation.com/files/447675/original/file-20220221-22-1ecl4d9.jpg?ixlib=rb-1.1.0&q=45&auto=format&w=754&h=1129&fit=crop&dpr=1 754w, https://images.theconversation.com/files/447675/original/file-20220221-22-1ecl4d9.jpg?ixlib=rb-1.1.0&q=30&auto=format&w=754&h=1129&fit=crop&dpr=2 1508w, https://images.theconversation.com/files/447675/original/file-20220221-22-1ecl4d9.jpg?ixlib=rb-1.1.0&q=15&auto=format&w=754&h=1129&fit=crop&dpr=3 2262w" sizes="(min-width: 1466px) 754px, (max-width: 599px) 100vw, (min-width: 600px) 600px, 237px"></a>
<figcaption>
<span class="caption">Women’s Work recognises the leadership of Indigenous women.</span>
<span class="attribution"><span class="source">Topbunk</span></span>
</figcaption>
</figure>
<p>Women’s Work celebrates the leadership of South Australian Aboriginal Women with striking black and white photographs. Tucked away down the hall on the second level is Fostering Ties, a series of images drawing attention to children in foster care. </p>
<p>This exhibition foregrounds invisibility as a way to contend with our own blind-spots, knowledge systems, biases and cultural frameworks.</p>
<p>What is invisible to us may not be to those from demographics, cultural or language groups that differ from ours. </p>
<p>Drawing attention to the invisible encourages us to shift our perspective. If we don’t have the answer to solve a problem, maybe another cultural perspective – or life form – does. </p>
<p><em>Invisiblity is at MOD until November 2022.</em></p><img src="https://counter.theconversation.com/content/176567/count.gif" alt="The Conversation" width="1" height="1" />
<p class="fine-print"><em><span>Kim Munro does not work for, consult, own shares in or receive funding from any company or organisation that would benefit from this article, and has disclosed no relevant affiliations beyond their academic appointment.</span></em></p>Disinformation, algorithms, big data, care work, climate change, and cultural knowledge can all be invisible. This exhibition brings them to the light.Kim Munro, Lecturer, University of South AustraliaLicensed as Creative Commons – attribution, no derivatives.tag:theconversation.com,2011:article/1775402022-02-21T19:07:44Z2022-02-21T19:07:44ZMandatory logins for ABC iview could open an intimate window onto your life<figure><img src="https://images.theconversation.com/files/447462/original/file-20220221-17-1nx6zk7.jpg?ixlib=rb-1.1.0&rect=5%2C0%2C1272%2C640&q=45&auto=format&w=496&fit=clip" /><figcaption><span class="caption">
</span> <span class="attribution"><span class="source">ABC</span></span></figcaption></figure><p>Last week, the ABC <a href="https://about.abc.net.au/press-releases/%E2%80%AFabc%E2%80%AFiview%E2%80%AFlogin-to-watch/">announced</a> it will begin to track the viewing habits of all users of its <a href="https://iview.abc.net.au">iview streaming platform</a> from March 15. This will be done by making users create an account and log in to watch shows and “benefit from the next stage of personalised services” such as “program recommendations [and] watchlists”.</p>
<p>The change was initially planned for the middle of last year, but was <a href="https://www.innovationaus.com/abc-quietly-delays-iview-login-plans-data-sharing/">delayed after heavy criticism</a> from privacy experts and others over the proposed arrangements for sharing and recording data. One point of contention was the ABC’s plans to share viewer data with Facebook and Google.</p>
<p>The ABC <a href="https://about.abc.net.au/statements/abc%E2%80%AFiview%E2%80%AFlogin-to-watch-faqs/">says</a> “significant work has been undertaken in providing effective privacy controls” during this delay. But nevertheless, <a href="https://www.salingerprivacy.com.au/2022/01/06/the-abcs-of-privacy/">critics maintain</a> the new provisions still involve sharing using data without full consent. </p>
<p>So how concerned should we be about our privacy here?</p>
<h2>All your data are belong to us</h2>
<p>For years we’ve known organisations such as Google and Facebook are collecting data on every search and social media post we make, and every website we visit. </p>
<p>Often the argument for collecting these data is similar to that used by the ABC: that collecting it provides for more personalised recommendations and a better user experience. However, tech companies also make billions using these data to sell personalised ads (and sometimes by selling the actual data).</p>
<hr>
<p>
<em>
<strong>
Read more:
<a href="https://theconversation.com/the-ugly-truth-tech-companies-are-tracking-and-misusing-our-data-and-theres-little-we-can-do-127444">The ugly truth: tech companies are tracking and misusing our data, and there's little we can do</a>
</strong>
</em>
</p>
<hr>
<p>They’re not the only ones keeping tabs on us. Loyalty cards such as Woolworths Everyday Rewards or Coles Flybuys do the same thing: tracking your purchases, adding them to a database, and mining them for information about your life. </p>
<p>If you buy 10 cans of cat food a fortnight, you probably have two cats. If you suddenly start buying 15, you’ve probably acquired a third. </p>
<p>Nappies, baby formula and baby food reveal how many kids you have, how old they are, and how they’re growing up. The ratio of Tim Tams to bread and milk can give clues as to your level of disposable income. </p>
<p>Despite this, millions of Australians scan these cards every day. It’s hard to know if they’ve fully weighed the pros and cons, or just never really thought about them.</p>
<h2>A healthy fear of your shadow (profile)</h2>
<p>So how much should we care about this? And how much do we? </p>
<p>When I put these questions to my students in an undergraduate class on Information Technology & Society, they mostly respond that if they’re doing nothing wrong then they have no reason to care if major corporations know what they eat for breakfast.</p>
<p>Older “mature age” students tend to feel differently, often raising concerns about what the data are used for, both now and potentially in the future. Older students may have had negative experiences with data, such as having a home loan disallowed over a credit report, while younger people may not look so far ahead.</p>
<figure class="align-center ">
<img alt="" src="https://images.theconversation.com/files/447473/original/file-20220221-14-8dly68.jpeg?ixlib=rb-1.1.0&q=45&auto=format&w=754&fit=clip" srcset="https://images.theconversation.com/files/447473/original/file-20220221-14-8dly68.jpeg?ixlib=rb-1.1.0&q=45&auto=format&w=600&h=510&fit=crop&dpr=1 600w, https://images.theconversation.com/files/447473/original/file-20220221-14-8dly68.jpeg?ixlib=rb-1.1.0&q=30&auto=format&w=600&h=510&fit=crop&dpr=2 1200w, https://images.theconversation.com/files/447473/original/file-20220221-14-8dly68.jpeg?ixlib=rb-1.1.0&q=15&auto=format&w=600&h=510&fit=crop&dpr=3 1800w, https://images.theconversation.com/files/447473/original/file-20220221-14-8dly68.jpeg?ixlib=rb-1.1.0&q=45&auto=format&w=754&h=640&fit=crop&dpr=1 754w, https://images.theconversation.com/files/447473/original/file-20220221-14-8dly68.jpeg?ixlib=rb-1.1.0&q=30&auto=format&w=754&h=640&fit=crop&dpr=2 1508w, https://images.theconversation.com/files/447473/original/file-20220221-14-8dly68.jpeg?ixlib=rb-1.1.0&q=15&auto=format&w=754&h=640&fit=crop&dpr=3 2262w" sizes="(min-width: 1466px) 754px, (max-width: 599px) 100vw, (min-width: 600px) 600px, 237px">
<figcaption>
<span class="caption">Data recorded today may be used for other purposes in the future.</span>
<span class="attribution"><span class="source">Shutterstock</span></span>
</figcaption>
</figure>
<p>Indeed, organisations like Electronic Frontiers Australia have argued this type of data collection can be <a href="https://www.efa.org.au/2014/11/06/ethics-big-data/">a slippery slope to profiling and bias</a>, with organisations using this to choose who should receive particular services or assistance. </p>
<p>The ever-growing collection of data comes at the same time as government moves to centralise their databases under the banner of <a href="https://my.gov.au">myGov</a>, tying all government services to Medicare or tax file numbers. </p>
<hr>
<p>
<em>
<strong>
Read more:
<a href="https://theconversation.com/is-chinas-social-credit-system-coming-to-australia-117095">Is China's social credit system coming to Australia?</a>
</strong>
</em>
</p>
<hr>
<p>We are still a long way from a dystopian situation like China’s social credit system, where all our behaviour feeds into a rating system that determines our access to services and housing, but these moves could make one easier to implement in future. </p>
<h2>How enjoying Q+A might raise tricky questions</h2>
<p>Which brings us back to the ABC and its plan to require every user to create a profile and log into its service. The main question here is the same one to ask when using a Flybuys card or creating a new social media account. </p>
<p>Does the convenience of sharing these data (with the ABC in this case), in terms of personal recommendations and watch lists, and indeed, the ability to access the service at all, balance what we think our data will ultimately be used for?</p>
<p>And when we ask this question, it helps to think in very broad terms. While in this case we’re just talking about viewing history and watch time, it’s not too dissimilar to cat food and nappies when you think about it. </p>
<hr>
<p>
<em>
<strong>
Read more:
<a href="https://theconversation.com/the-privacy-paradox-we-claim-we-care-about-our-data-so-why-dont-our-actions-match-143354">The privacy paradox: we claim we care about our data, so why don't our actions match?</a>
</strong>
</em>
</p>
<hr>
<p>Significant amounts of information could be inferred from our viewing habits: everything from our political leanings to our attention span. What that can then be used for is anyone’s guess.</p>
<p>That’s not to say you shouldn’t create an account, but rather that you need to go in with your eyes wide open. Think about what iview means to you, what data might be shared, and how it might be used. And then decide if you really love Bluey all that much after all.</p><img src="https://counter.theconversation.com/content/177540/count.gif" alt="The Conversation" width="1" height="1" />
<p class="fine-print"><em><span>Michael Cowling does not work for, consult, own shares in or receive funding from any company or organisation that would benefit from this article, and has disclosed no relevant affiliations beyond their academic appointment.</span></em></p>The ABC’s decision to force viewers to create accounts to watch shows online raises concerns over privacy.Michael Cowling, Associate Professor – Information & Communication Technology (ICT), CQUniversity AustraliaLicensed as Creative Commons – attribution, no derivatives.