Big Data and Drones
Since 2009, the US has carried out more than 500 covert drone strikes in Pakistan, Yemen and Somalia, killing thousands it judged posed a terrorist threat. The U.S. Administration has hailed drones as “our most precise weapon to date”, a weapon that allows for accurate strikes with little collateral damage and zero risk to the soldiers flying them. Yet over and over again, mistakes seem to be made. An imam dies in a drone strike in Yemen days after preaching against Al-Qaeda. A wedding is ruined when a drone attack kills a dozen guests. A community meeting of elders gathered to resolve a mining dispute in Pakistan ends in disaster. And despite investigations by journalists and human rights organisations, the mistakes persist. This talk will examine the role of metadata in the US drone programme and how an overreliance on so-called “big data” results in life and death consequences for communities thousands of miles away.
Jennifer Gibson is a US-trained lawyer and leads Reprieve’s Pakistan drones work. In 2012, she led a delegation of Pakistani drone victims to the US, where for the first time US Congressmen and women heard testimony from those directly affected by drones. She has authored or co-authored a number of reports on drones, including You Never Die Twice: Multiple Kills in the US Drone Programme, and the Stanford/NYU study, Living Under Drones. She recently served on the University of Birmingham’s Policy Commission on the Security Impact of Drones, chaired by Sir David Ormand.
Contact: Jennifer.email@example.com or on Twitter @jennifermgibson
Data analytics and the UK elections
This talk will look at a joint Media Standards Trust / King’s College London Big Data analysis of candidates and political influencers on Twitter over the course of the six week UK General Election campaign 2015. Looking at political engagement on Twitter, this partly qualitative and largely quantitative analysis has sought to trace who has set the agenda for the 2015 UK General Election. With this in mind, the talk will put Twitter in the context of our examination of Mainstream News Media, Party Political Agendas and Public Opinion. By introducing our software analysis tool – called Steno – the talk will outline the distinctive methodology we adopted, suggesting some alternatives for Big Data analysis. By differing in its approach from similar studies, which have either concentrated on sentiment analysis, network analysis, or the dynamics of social media, this study suggests a more focused analysis of Twitter profiles and content. There will be reference to some initial draft findings, though given the analysis runs to the eve of the conference these will necessarily only be indicative.
Martin Moore is director of the Media Standards Trust and a visiting senior research fellow at King’s College London. He is the author of ‘The Origins of Modern Spin’ and has been doing media research within and outside academia for almost two decades.
What does the politics of big data look like if big data is snake oil?
It is tempting to take the claims made for big data by its proponents at face value, and use these claims as a basis for assessing its potential political impacts, upon modes of politics, state and private surveillance capacities and even upon political subjectivity. Starting from the basis that critical researchers should be cautious about the rhetorical claims made for big data, and using analogies from other areas of surveillance research, this talk will sketch out what a politics of big data might be, if the claims of big data advocates are approached skeptically. It will offer some signs that a particular big data claim might be suspect, and also offer some reflections on what it is about our contemporary politics that provides for the acceptance of grand claims about the capabilities of big data.
David Barnard-Wills is a Political Scientist with an interest in the the politics of surveillance, identity, technology and security. Senior Research Analyst at Trilateral Research
Digital Transparency and The Politics of Open Data
In recent years the concept of open data has developed from being a niche idea at the margins of software development communities to playing a central role in global information policy. This paper draws on a combination of historical and empirical research to examine open data as a contested political concept that is continually reconfigured in response to shifting ideals, conceptions and practices of governance and democracy in different contexts. This includes work towards a “genealogy of open data”, as well as the findings from several research projects at the Digital Methods Initiative to map the politics of open data as an issue on digital media. It concludes with reflections on open data initiatives as sociotechnical assemblages and on emerging forms of intervention calling not just for the disclosure of information but for more fundamental changes in the composition of information infrastructures that organise collective life.
Jonathan Gray is Doctoral Researcher in the Department of Politics and International Relations at Royal Holloway (University of London), Research Associate at the Digital Methods Initiative (University of Amsterdam); and Research Fellow at the Tow Center for Digital Journalism (Columbia University).He is also Director of Policy and Research at Open Knowledge, a global civil society organisation dedicated to opening up public information, research and culture to benefit the lives of citizens around the world. More about him can be found at jonathangray.org and he is on Twitter at @jwyg.
Social Media Research: Bringing Big Data Down to a Manageable and Meaningful Scale
In this presentation I outline several ways in which the issue of platformativity, i.e. the structuring of social action and interaction by social network services, may begin to be offset with multiple datasets and mixed-methods. The significance of this problem has become apparent in recent studies highlighting the commercial logic and algorithmic techniques whereby popular services such as Facebook and Twitter control the access of independent researchers to user-generated data. In my talk, I suggest potential approaches to salvaging social media research data. My main proposition is that tried and emergent methods of data validation may aid with grasping the scope and intricate nature of the intersection between technological design and user agency with the aim to disentangle ensuing implications for social action such as political protests.
Dr Dan Mercea received his PhD in communication studies from the Department of Sociology, University of York. Before the completion of his doctorate he became Teaching Fellow in Political Sociology at York and from September 2011 to September 2013 he was Senior Lecturer in Politics at The Hague University of Applied Sciences in the Netherlands. During that time he was also Visiting Lecturer in Political Communication at the Catholic University of Lille, France where he continues to be Associate Research Fellow.
The potential and limit of social media for studying elections
Dhiraj Murthy, Goldsmiths, University of London
Social media have been prominently covered in the press for their perceived role in activism, disaster recovery, and elections amongst other things. Given we are in the midst of the UK General Election, this talk explores the potential and limit of understanding elections via Big Data derived from social media. In the case of elections, Twitter, Facebook, Instagram, Tumblr and other social media have been used actively by candidates and voters alike in a diverse range of elections around the world. However, not only have social media often been found to be a poor predictor of electoral success, but they are often heavily biased sociopolitical spaces. This talk argues that social media is often paraded as an indicator of a political ‘pulse’. However, many social media are reactive in the context of elections rather than predictive. Additionally, there are major limitations to applying sentiment analysis and other machine learning methods to election-related social media data. Also, there is a belief that social media can easily reveal public sentiment towards candidates, serving as a replacement to polling. However, most often social media such as Twitter are neutral towards candidates. Explanations are provided as to what potential and limits these data have towards understanding political and social phenomena more broadly.
Dhiraj Murthy is Reader of Sociology at Goldsmiths College, London. His research focus on Sociology, Research Methodology, and Social Media.
The Performance Platform: Governing through Design
The routine operations of government are being transformed through the implementation of new online platforms, web analytic and user research techniques, ‘agile’ (management) methods, and the introduction of design principles. Government is understood as a standardized platform; its services subject to continuous performance measures and data analysis, and its citizens reconfigured as users. Recognising that data-driven systems have long been part of business intelligence and performance management, I reflect upon an ongoing study of Government Digital Service to consider what is unique about these practices coming together in public administration.
Nathaniel Tkacz is Assistant Professor in CIM, The University of Warwick. He is currently PI on the ESRC-funded ‘Interrogating the Dashboard: Data, Indicators and Decision-making’. His new book Wikipedia and The Politics of Openness (The University of Chicago Press, 2015) was selected as THE’s first book of the week for 2015. Other recent publications include The MoneyLab Reader (2015) and Digital Light (2015).
What counts in social media research?
While algorithmically calculated popularity rankings influence scores or relevance measures have faced increased critical inquiry, little attention has been paid to the composition of first order metrics, such as counts of shares, tweets or hashtags. But what do we actually count when we aggregate tweets, @replies, likes or shares? Metrics, the presentation suggests, need to be seen as epistemic devices, informed by diverse use cultures and platform politics. Based on empirical work on a 1% Twitter sample, the presentation suggests that metrics do not count but calculate and have to be considered as lively – animated by users, their practices and mediating devices.
Carolin Gerlitz is Assistant Professor in New Media & Digital Culture, member of the Digital Methods Initiative and conducts research on digital culture and methods.
Chrono–Engineering the Interface: Network Latency and Deliberative Devices
This presentation examines commercial performance strategies for user-experience (UX) design as cultural techniques of precognitive capitalism. I focus, in particular, on network latency and speed optimization in service of conversion rates or measurable user events. This typically involves a process of chrono–engineering the interface through the management of responsive design, the optimization of loading paths and monitoring of usage patterns. Weighing such dynamics against debates in affect theory on the technicities and temporalities of non-conscious thought, I consider how hetero-chronic characteristics of information transfer are effectively managed by designers and developers, and speculate on how these methods might be reworked for more deliberative experiences.
Michael Dieter is Assistant Professor in CIM, The University of Warwick. His work has appeared in the journals M/C, Fibreculture, differences and the Australian Humanities Review. With David M. Berry, he is the co-editor of the forthcoming collection Postdigital Aesthetics: Art, Computation and Design (2015).
Hacking the Social Life of Big Data
It is paradoxical that questions of agency arise in relation to big data, considering that collectively, we are a leading source of its generation. Data literacy, as conceptualized in the AHRC grant: ‘Our Data Ourselves’, will be put forward as a critical emergent field that is aiming to develop a set of competencies and knowledge to empower people to understand the dynamic flows and processes related to our steadily growing digital footprint. Key aspects of the digital ecosystems that facilitate the capture of big data on our mobile devices will be examined in relation to how gaining access to one’s own data, not only augments the agency of the individual but of the collective user.
Jennifer Pybus is Senior Lecturer in the London College of Communication at University of the Arts London.
Visual Cultures of Big Data
The recent rise in social media platforms and apps focused on image sharing highlights a series of important issues in terms of the politics of big data. How do we deal with this recent explosion in image data? Images don’t lend themselves easily to computational techniques, which are unable to get at their meaning in any significant way. This presentation will highlight work from the Visual Social Media Lab, which works towards developing methodological and theoretical strategies for the capture and interpretation of social media image data at different scales. This work also draws attention to the need to consider wider cultures of visibility. For example: how do we get to see these images in our feeds and streams at all? Do we see them as images? This presentation therefore also highlights the need to interrogate these new regimes of algorithmic visibility, including questions around accountability.
Farida Vis is a Faculty Research Fellow in the Information School, at the University of Sheffield. She is the Director of the Visual Social Media Lab.
Inside the data activism ecology
With the diffusion of ‘big data’, citizens become increasingly aware of the critical role of information in modern societies. This awareness gives rise to socio-technical practices rooted in technology and data, which I term ‘data activism’. Data activism has a critical approach to big data at its core, and takes two forms: firstly, citizens resist the threats to civil rights that derive from corporate and government privacy intrusion (‘re-active data activism’), and, secondly, they take advantage of the possibilities for advocacy offered by big data (‘pro-active data activism’). Data activism emerges from the open-source and hacker movements, but overcomes their elitist character to involve ordinary users. It appropriates and subverts technological innovation and software, their language and epistemological approach, for social change purposes. This talk explores the data activism ecology, sketching a potential approach to the study of the politics of big data from a civil society point of view.
Stefania Milan is Assistant Professor of New Media and Digital Culture at the University of Amsterdam, and a research associate at the Tilburg Institute for Law, Technology and Society (Tilburg University) and the Internet Policy Observatory of Annenberg School of Communication (University of Pennsylvania). Her research explores grassroots internet activism, cyberspace governance, technology and participation, and data epistemologies. In December 2014 she has been awarded a Starting Grant of the European Research Council for the project ‘Data activism: The politics of big data according to civil society’. Stefania holds a PhD in political and social sciences of the European University Institute, Italy. Prior to joining the University of Amsterdam, she worked at the Citizen Lab (University of Toronto) and Tilburg University. She is the author of Social Movements and Their Technologies: Wiring Social Change (Palgrave Macmillan, 2013), and co-author of Media/Society (Sage, 2011), and her work has appeared in a variety of academic journals.
The Dark Side of the Quantified Self
Phoebe Moore (w Lukasz Piwek)
This paper (Moore and Robinson 2015) looks at the recent rise in the use of wearable sensory technological devices, or wearable devices and other self-tracking technologies (WSTT). The technologies include devices linked to smartphone software apps. WSTT measure and track arousal and performance both mental and physical via accelerometers, Bluetooth, triangulation algorithms and infrared sensors. I have broken down categories in which WSTT is used, firstly in fashion and garments; secondly in health and personal fitness; and the third, and increasingly of interest to investors, in workplaces. The workplace usage interests me the most because more than 13 million fitness tracking devices will be incorporated into workplace wellbeing programmes from 2014 – 2019 according to ABI research. But I have identified some issues in integration of WSTT into workplaces in its early iterations. The ‘dark side’ of the quantified self (Moore and Piwek 2015) involves corporate profiteering and lack of legal regulation and the ways that the technologies are being used in different sectors. In order to theorise the implications for this movement effectively we claim that a reinvestigation of the dominant ontology of dualism, with recognition of the exploitative potentials for this framing, as well as interrogation of Marxist, post-Marxist and new materialist exposition, is required.
Phoebe Moore is a Senior Lecturer at University of Middlesex London in the Law and Politics department. Trained as a social scientist, Phoebe writes about the worlds of work and labour including the impact of technology and changing regulation within the global neoliberal political economy. Her third in-process monograph, with Routledge, is entitled The Quantified Self at Work. Among other things, she has published on corporeal capitalism and the quantified self (Edward Elgar 2014), peer to peer production (Journal of Fibreculture and Capital & Class), labour struggle in East Asia (Palgrave and IBTauris), and the International Labour Organisation and ‘decent work’ (Globalizations, Pluto, Global Labour Journal fc).
Big Data and the Paradox of Diversity
In this talk, I will argue that at the core of “big data” as a politically charged phenomenon lies our increasing capacity to detect and act on variation in a quickly growing number of domains. Fine-grained differentiation – between people, things, ideas or situations – has historically been difficult and costly. The availability of both data and the techniques to mine them makes it feasible to individualize far beyond the granularity of postal codes, income brackets or skin color. Machine learning, in particular, makes it possible to relate any variable to a desired outcome (e.g. a low rate of credit defaulting), transforming inequality, in the broadest sense, into economic opportunity and leading to highly problematic outcomes. While notions like surveillance or discrimination capture part of the problem, I will argue that we need new concepts to fully account for the challenges that the interested ad-hoc groupings big data make possible pose to liberal democracy.
Bernhard Rieder is Associate Professor in New Media and Digital Culture at the University of Amsterdam and a collaborator with the Digital Methods Initiative. His research focuses on the history, theory, and politics of software and in particular on the role advanced information processing techniques play in the production of knowledge, culture, and sociality. He has written numerous research tools, in particular for the study of the social web, and is currently working on a book on the conceptual underpinnings and political challenges of algorithmic ordering.
Tracking productive subjects: corporate wellness programmes, self-tracking and control through data
This paper will explore the formation of an affective socio-technical assemblage through the use of digital self-tracking devices in corporate wellness programs. Corporate wellness initiatives conflate the health, fitness and wellbeing of individuals with the productivity of the organisation. Increasingly such programs are using digital apps and devices to increase employee engagement. Analysis of interviews with managers involved in the application of such initiatives, and a discourse analysis of related literature, will be presented. This will show that the key focus of such programs is the development and maintenance of affective relationships; employers demonstrate that they care through data. This caring is expressed by enabling employees to quantify and analyse their exercise and providing them with new ways to be better, healthier, more productive employees. This analysis draws on Actor Network Theory and Boltanski and Chiappello’s The New Spirit of Capitalism to show how this socio-technical assemblage is formed.
Chris Till is a senior lecturer in Sociology in the School of Social, Psychological and Communication Sciences at Leeds Beckett University. He is a sociologist of health with a particular focus on bodies and technologies. His current work is focused on the impact of data, and particularly digital self-tracking, on how corporate power and control are exercised through bodies and affective relationships. Recently he has published the article ‘Exercise as Labour: Quantified Self and the Transformation of Exercise into Labour’ in the journal Societies. He tweets at @chrishtill and keeps the blog This is Not a Sociology Blog.
Digital Vigilantism: Visibility as a Weapon
This paper considers a research agenda for the study of digital vigilantism as a user-led violation of privacy that not only transcends online/offline distinctions, but also complicates relations of visibility and control between police and the public. Digital vigilantism (DV) is a process where citizens are collectively offended by other citizen activity, and coordinate retaliation on mobile devices and social platforms. The offending acts range from mild breaches of social protocol to terrorist acts and participation in riots. The vigilantism includes, but is not limited to a ‘naming and shaming’ type of visibility, where the target’s home address, work details and other highly sensitive details are published on a public site (‘doxing’), followed by online and in-person harassment. The visibility produced through DV is unwanted (the target is typically not soliciting publicity), intense (content like text, photos and videos can circulate to millions of users within a few days), and enduring (the vigilantism campaign may be top search item linked to the target, and even become a cultural reference).
Daniel Trottier is an Assistant Professor in the Department of Media and Communication at Erasmus University Rotterdam. His current research considers the use of social media by police and intelligence agencies, as well as other forms of policing that occur on these platforms. As part of this research, he has participated in two European Commission projects on security, privacy and digital media. Daniel has authored several articles in peer-reviewed journals on this and other topics, as well as Social Media as Surveillance with Ashgate in 2012, Identity Problems in the Facebook Era with Routledge in 2013, and Social Media, Politics and the State (co-edited with Christian Fuchs) with Routledge in 2014.
Blow up the Pattern: On the Design Of ‘humanity’ in the Age of Big Data
This presentation will discuss how ‘Big Data’ shapes our contemporary discourse beyond computing. For this, it will look at the effects ‘Big Data’ has on the concept of ‘humanity’ from two different perspectives. Analysing visualisations and interfaces, it will critically evaluate the idea that by looking through unstructured data we get a glimpse of what this could be: ‘humanity’. Having problematized this fascination, it will then radically change its perspective to ask how an individual human can make himself/herself heard in the midst of such a ‘big data’ stream – or how to blow up the pattern.
Mercedes Bunz is Senior Lecturer at the University of Westminster, London, where she teaches Digital Media and Journalism. She is also a member of the interdisciplinary network for the critical humanities terracritica.net, and has been the technology reporter of The Guardian. She writes about technology, media, critical theory and journalism. Her last book is The Silent Revolution: How Algorithms Changed Knowledge, Work, Journalism, and Politics Without Making Too Much Noise (Palgrave Macmillan 2014).
Will big data cure us? Participation and surveillance in medicine
The idea that medicine should be tailored more closely to the specific characteristics of individual patients has gained currency in recent decades. Since the Human Genome Project, whose clinical benefits have generally been considered underwhelming, more and more supporters of such “personalised” or “precision medicine” call for the consideration of wider ranges of data, including non-molecular, clinical and lifestyle-related data, in clinical decision making. Data-driven medicine is seen as a catalyst for a more personal and precise (and as some argue, more cost-effective) medicine. A group of scientists in Boston even went as far as suggesting that our credit card purchases, our social media activity, and our criminal records should be integrated along with our genome sequence and your other health records to ‘personalise’ our healthcare. In my talk I would like to draw attention to the many ways in which contributions from patients and health people (“patient work”, as Anselm Strauss called it) are a presupposed, required and facilitated in this endeavour. People become passive data sources and active participants in their own surveillance at the same time. The notion of the “paying volunteer” in citizen science projects illustrates this phenomenon. I will end with a proposal to rethink some of the very concepts and terms that we use to frame and phrase the problem of participatory surveillance in the health domain. The paper can be found here http://www.ncbi.nlm.nih.gov/pmc/articles/PMC4114418/
Dr Barbara Prainsack is a Professor in Sociology at the Department of Social Science, Health & Medicine. Her main research interest lies in the interface between science and regulation, as well as in how the relationship between the two is related to individual and collective identities.