Search Results

Data analytics have emerged in recent years as a family of overlapping, competing and hybridising products and practices. They have been championed by technology companies, academics, business users and governments alike, and in a short period of time have earned business developers and adopters billions of pounds in revenue and unprecedented levels of market domination. Data analytics have also provided distinct benefits in terms of an increasing democratisation of digital tools, but at the same time are giving rise to increasing levels of societal and governmental concern. This chapter has four aims: to help intelligent outsiders and old school data analysts make sense of the many competing methodologies and technologies that inhabit the data analytics ecosystem; to assist readers understand which of the many techniques and methodologies represent genuine additions to the state of the art rather than simply old wine in new bottles; to provide a brief overview of the software tools currently available for data analytics; and to identify societal issues and concerns that attend this family of technical and social practices, and the extent to which they are being adequately addressed by developers, users and society at large.

Policy analysis relies on data collected at discrete intervals along the policy cycle, from problem identification through evaluation. Policy analytics, in contrast, represents the combination of new, ubiquitous, and continuous data sources—from Internet search and social media to mobile smartphones, Internet of Everything (IoE) devices, and electronic transaction cards—with new data analytics techniques for informing and directing policy choices. New technology platforms also offer the possibility of small-scale policy experiments that can be piloted with their effects precisely observed in real-time. This big data + analytics + real-time experiments approach offers a significant change to the traditional practice of policy analysis. This chapter describes the movement from policy analysis to policy analytics, discusses emergent examples and potential applications, and concludes with questions that can guide the appropriate adoption of policy analytics for supporting policymaking.

This chapter highlights how the police resist and contest big data analytics. Their resistance stems in large part from the proliferation of high-frequency observations and data collection sensors resulting in the police themselves coming under increased surveillance. New developments in surveillance technologies are frequently viewed with suspicion, with officers believing technology is a means of deskilling, entrenching managerial control, devaluing experiential knowledge, and threatening their professional autonomy. Novel tech ultimately serves to reinforce old divisions—such as those between managers and patrol officers—even as it creates new distinctions within the Los Angeles Police Department (LAPD). Understanding these patterns of contestation underscores how big data is ultimately social. It also allows one to consider the extent to which data-based surveillance is—and is not—associated with deeper organizational change.

Advanced analytical methodologies and data visualizations are transforming how human resource data are used in organizations. Through the application of these methods, 360 Feedback can be enhanced; conversely, a variety of talent management objectives can be enhanced by including 360 Feedback. This chapter discusses how analytics can help support organizations and individuals as they tackle these systematic questions. The chapter begins with an overview of major talent management functions (selection, promotion, learning and development, succession and workforce planning, analytics for measuring progress), showing how each can incorporate 360 Feedback data. Attention is devoted to multirater feedback as a target of inquiry itself, discussing how new analytic approaches (unstructured data, linkage analysis, network analysis, reporting, and visualization) can enhance traditional practice. We close with a discussion of key considerations about foundational issues such as data privacy.

In this chapter, Garnett and Hughes focus on the role of big data in accessing information from public inquiries. Looking at the Chelsea Manning court martial in the US and the Leveson Inquiry in the UK, they argue that the manner in which information pertaining to inquiries is made public is, at best, unsatisfactory. They propose a variety of means to make this information more accessible and hence more transparent to the public through employing big data techniques.

What is big data, and what does it have to do with the humanities? The Snowden revelations have drawn attention to the opportunities and dangers to the gathering of large collections of data, including the collecting of text messages and email. Techniques that digital humanists have used in the study of individual texts are now being scaled up to study large collections. The digital humanities have a valuable historical and ethical perspective on big data analytics. Questions about what to do with too much information go back to Plato. Questions about the completeness of data, the usefulness of metadata, and the value of analytics can help us understand what big data can and cannot do. In particular we need to be careful of false positives, or false predictions based on data too large to check with other methods.

This chapter traces the history of quantification in policing, from pin maps to predictive algorithms. It examines the surveillance landscape, starting with the “scientific turn” in policing in the early 20th century, then moving to the rise of evidence-based policing, the Information Sharing Environment (ISE) that emerged after the terrorist attacks of 9/11, and then predictive algorithms and big data analytics put to work in modern policing. Historically, the police collected most of the information they use in the course of their daily operations themselves. However, the chapter highlights the growing role of the private sector—for data collection and the provision of analytic platforms—in policing. Both the past and present of policing are highly racialized, so it also describes how data is positioned as an antidote to racism and bias in policing. The chapter concludes with an overview of data use and technologies at work in the Los Angeles Police Department (LAPD).

This chapter introduces the book through highlighting recent cases of ethical and democratic concern arising through the use of data analytics (so-called “big data”). The aftermath of the election of Donald Trump and the Brexit referendum were dogged by claims that the elections had been manipulated through microtargeting by companies such as Cambridge Analytica in favour of the winning parties. The Introduction concludes with an overview of each following chapter in the book.

The quest for data-driven campaigning in 2012—creating massive databases of voter information for more effective micro-targeting—found greater efficacy and new controversy in 2016. The Trump campaign capitalized on the power of digital advertising to reach the public to engage in unprecedented mass-targeted campaigning. His campaign spent substantially more on Facebook and other digital media paid ads than Clinton. Yet, the company that Trump worked with, Cambridge Analytica, closed up shop in 2018 under a cloud of controversy about corrupt officials and voter manipulation in several countries, as well as ill-begotten data of Facebook users that drove their micro-targeting practices. The Clinton campaign modeled itself on data-driven successes of the Obama campaign, yet the algorithms that drove their decision making were flawed, thereby leading her campaign to underperform in essential swing states. Similar to the Romney campaign’s Narwhal challenges on Election Day when the campaign effectively was flying blind on get-out-the-vote numbers, the Clinton plane was flying on bad coordinates, ultimately causing her campaign to crash in critical swing states.

One of the most prominent features of mass digitization is the automated processing of works for various research-related and commercial purposes. This includes text mining or linguistic analysis over masses of texts, image analysis, information extraction, automatic translation, data mining for behavioural profiling, and so on. With a few exceptions, copying for automated processing or computational analysis does not feature in statutory language and its status remains uncertain. While, historically, ‘machine-reading’ has challenged copyright norms at some instances, there is currently need for a careful consideration of the parameters of its permissibility.

This chapter details a conversation between open data advocates and a civil servant in charge of the process, which reveals the challenges of getting government data made open. Without an injection of funds, an open data initiative called the Regional Data Lab, was in danger of winding down. Government had very little interest in making their data available, and even less enthusiasm for spending money during austerity. And open data was not free data; somebody had to pay for the labour of preparing data for release and building the necessary data infrastructure. What the Regional Data Lab does is take what data are already openly available and make them useable for those that lack the skills to build their own tools so they can use them in formulating policy. Rather than negotiating separate contracts every time, it would make more sense to simply centrally fund the Regional Data Lab to provide a suite of core data services. However, the advocates are more interested in the development of a national open data repository and access to more data, and a coordinated approach to providing data analytics for the public sector.

This chapter looks at the promise and peril of police use of big data analytics for inequality. On the one hand, big data analytics may be a means by which to ameliorate persistent inequalities in policing. Data can be used to “police the police” and replace unparticularized suspicion of racial minorities and human exaggeration of patterns with less biased predictions of risk. On the other hand, data-intensive police surveillance practices are implicated in the reproduction of inequality in at least four ways: by deepening the surveillance of individuals already under suspicion, codifying a secondary surveillance network of individuals with no direct police contact, widening the criminal justice dragnet unequally, and leading people to avoid institutions that collect data and are fundamental to social integration. Crucially, as currently implemented, “data-driven” decision-making techwashes, both obscuring and amplifying social inequalities under a patina of objectivity.

A substantial body of copyright infringement defences is primarily available to institutional users, such as educational establishments, libraries, and archives. In light of the advent of the Internet and mass digitization, the availability of defences has been enlarged through a set of legislative instruments, such as the Orphan Works Directive and the Directive on Copyright in the Digital Single Market. Public policy privileges are meant to make allowances for modern methods of teaching provision, such as online courses, distance learning, and cross-border education programmes, as well as exempt from infringement new methods of carrying out research, such as text mining and data analytics, and enable value extraction from the plethora of works that are currently out of print. Although the policy reason behind the expansion of available defences has been the promotion of growth in the educational and cultural sector, there is a strong public interest underpinning the very presence of these exceptions in the statute. This has to do with the promotion of a rigorous public domain, whereby certain works shall be made more accessible for users to use and re-use. Subject to examination in this chapter is the breadth of these permissible activities and their ability to accommodate modern online services, including also the defences available for uses made by public administration.

Stelzer and Veljanova argue in this chapter for a new ethical compass with which to approach big data concerns, such as privacy, equality, and discrimination. They identify key ethical concerns which often arise in cases regarding big data and then provide a framework through which we might approach these concerns such that we can have a degree of certainty that we have not overlooked ethical worries.

This chapter reveals how choices and decisions concerning the analytics applied to data shapes outcomes, through an account of a working session between academics and a government minister to devise and implement an 'objective' method for allocating government funding. The nub of the problem was the Minister had a very particular outcome in mind. He wanted the investment from his new scheme to be spread across as many constituencies as possible, and certainly the ones that traditionally voted for his party or those that might swing away from the government. However, he did not want to be seen to allocate the funding on political grounds, nor run the scheme on a competitive basis. Instead, he wanted to be able to say that the monies had been apportioned using a statistical formula that assessed need objectively. Creating a formula for producing a map that pleased the Minister proved to be trickier than anticipated. In part, this was because he had his own ideas about which variables were good indicators of relative deprivation and need.

MacWillie’s chapter develops an ontological understanding of the infrastructure underlying big data applications through an historical overview of developments in information communications technology since the 1950s. This leads him to conclude that big data is a fundamentally new object in the world, bringing with it key issues of richness and complexity in computer networks.

The principles of data protection should apply to any information concerning an identified or identifiable natural person. Personal data which have undergone pseudonymisation, which could be attributed to a natural person by the use of additional information should be considered to be information on an identifiable natural person. To determine whether a natural person is identifiable, account should be taken of all the means reasonably likely to be used, such as singling out, either by the controller or by another person to identify the natural person directly or indirectly.

The 2012 presidential candidates refined proven practices and ran the most data-driven campaigns in history. The candidates deployed social media, strategic online ad buys, and used their websites as the cornerstones of their campaign practices in the increasingly complex, hybrid media environment. Obama’s and Romney’s campaigns produced a variety of tactics to interact with supporters in a way that suggested that controlled interactivity had been perfected. They built massive voter files to target usual demographic groups while expanding to new groups typically unreached by campaigns and conducted careful message testing to yield maximum effect. Yet, for the carefully scripted work to structure interactivity between supporters and the campaign and among supporters to greatest advantage for the candidate, a substantial challenge remained: how to manage messaging in the complex, hybrid media environment where gaffes and opposition discourse can be amplified in ways unintended and with unknown consequences for campaigns.