Helen Kennedy. Post, Mine, Repeat. Social Media Data Mining Becomes Ordinary

Transcription

1 Post, Mine, Repeat

2

3 Helen Kennedy Post, Mine, Repeat Social Media Data Mining Becomes Ordinary

4 Helen Kennedy Department of Sociological Studies University of Sheffield Sheffield, UK ISBN DOI / ISBN (ebook) Library of Congress Control Number: The Editor(s) (if applicable) and The Author(s) 2016 The author(s) has/have asserted their right(s) to be identified as the author(s) of this work in accordance with the Copyright, Designs and Patents Act This work is subject to copyright. All rights are solely and exclusively licensed by the Publisher, whether the whole or part of the material is concerned, specifically the rights of translation, reprinting, reuse of illustrations, recitation, broadcasting, reproduction on microfilms or in any other physical way, and transmission or information storage and retrieval, electronic adaptation, computer software, or by similar or dissimilar methodology now known or hereafter developed. The use of general descriptive names, registered names, trademarks, service marks, etc. in this publication does not imply, even in the absence of a specific statement, that such names are exempt from the relevant protective laws and regulations and therefore free for general use. The publisher, the authors and the editors are safe to assume that the advice and information in this book are believed to be true and accurate at the date of publication. Neither the publisher nor the authors or the editors give a warranty, express or implied, with respect to the material contained herein or for any errors or omissions that may have been made. Printed on acid-free paper This Palgrave Macmillan imprint is published by Springer Nature The registered company is Macmillan Publishers Ltd. London

5 ACKNOWLEDGEMENTS Much of the research discussed in this book received funding from UK Research Councils and from other sources, and for this I m very grateful. Thanks to funders as follows: The Arts and Humanities Research Council (AHRC), which supported Understanding Social Media Monitoring (Grant AH/ L003775) with a Research Fellowship. This funding made possible the research discussed in Chapter 6, some of the research discussed in Chapter 8, and the writing of this book. The Engineering and Physical Sciences Research Council (EPSRC), which supported Digital Data Analysis, public engagement and the social life of methods (Grant ) and Public Engagement and Cultures of Expertise (Grant EPSRC-CCN2013-LEEDS), discussed in Chapter 4. Various small grants from the University of Leeds, including: HEIF V (Higher Education Innovation Fund, Reference ) IGNITE (References and ) The Creative and Cultural Industries Exchange (Reference ). These grants funded interviews in social media data mining companies, discussed in Chapter 5, and focus groups with social media users, discussed in Chapter 7. v

6 vi ACKNOWLEDGEMENTS Some of the material in this book is derived in part from journal articles, although it has all been cut, edited, re-worked and re-purposed for inclusion here, so it may no longer be recognisable as originating in these spaces. I m grateful to the publishers of the following journals (and articles) for permission to develop this material here: Kennedy, H. (2012) Perspectives on sentiment analysis, Journal of Broadcasting and Electronic Media, 56(4): (Chapter 5), copyright Taylor & Francis/Society, available online at: tandfonline.com/doi/pdf/ / Kennedy, H., Elgesem, D. and Miguel, C. (2015) On fairness: user perspectives on social media data mining, Convergence: the international journal of research into new media technologies 21(4): PAGES (Chapter 7), copyright Sage Publications, available online at: doi: / Kennedy, H., Moss, G., Birchall, C. and Moshonas, S. (2015) Balancing the potential and problems of digital data through action research: methodological reflections, Information Communication and Society 18(2): (Chapter 4), copyright Taylor & Francis/Society, available online at: doi/abs/ / x ?journalcode=rics20#. Vel0Y854jGt My AHRC Fellowship funded research visits to some of the best places in the world for thinking about the issues this book addresses. I m very grateful to the following people for arranging visits to the following places: Principal Researcher Nancy Baym, Microsoft Research Social Media Collective, spring Professor José van Dijck, Comparative Media Studies, University of Amsterdam, summer Professor Helen Nissenbaum, Media, Culture and Communication, New York University, spring I m extremely thankful that so many people, including social media users and employees in social media data mining companies, in public sector and in other organisations, have been willing to spend time talking to me and my researchers about their experiences with mining social data. Without them this book would not have been possible. Another group of

7 ACKNOWLEDGEMENTS vii people without whom this book wouldn t exist are the many collaborators, post-doctoral research fellows and research assistants with whom I have worked and who have contributed to my thinking and to the words on these pages. These include: Christopher Birchall, Stephen Coleman, Ivar Dusseljee, Dag Elgesem, Julie Firmstone, Jan-Japp Heine, James Hennessey, Patrick McKeown, Cristina Miguel, Stylianos Moshonas, Giles Moss, Matthew Murphy, Katy Parry, Christoph Raetzsch, Stuart Russell, Helen Thornham and Nancy Thumim. Thanks to all of you. During the time I ve been writing this book, there have been many exchanges, presentations, seminars, workshops and conferences around the world which have helped me to develop my ideas. I m especially grateful to these people for stimulating conversations, online and off: David Beer, David Hesmondhalgh, Noortje Marres, Thomas Poell, José van Dijck and the many brilliant thinkers at the Digital Methods Initiative at the University of Amsterdam and at the Microsoft Research Social Media Collective in Boston, USA. And a big thank you to all the folks at the Universities of Leeds and Sheffield, my former and current employers, two great universities in two great Yorkshire cities, who helped me finish this book in all sorts of ways. Everybody knows some Daves, and if you re lucky, you know some good ones. I do. David Hesmondhalgh (always a supportive, critical friend) and David Beer took the time to read a draft of this manuscript and to provide really helpful comments, and I m very grateful to both them for that. My favourite Dave, though, is the one I live with. He drew the picture on the front of this book with a pencil and a piece of paper no datafication there and he learnt a bit about data in society in the process. He puts up with me and provides the jokes. Thanks Dave. Friends and family make writing books bearable, and I m grateful to every one of them who bothered to say how s the book?, especially those who said it repeatedly. This, my friends, is how the book is. The things that interest my kids are wonderfully different from the things that this book is about, and that more than anything keeps me happy and sane, so thank you for being yourselves Pearl and Lonnie. I would wait excitedly with you for the delivery of a packet of loom bands or a special edition videogame book forever. I end with an acknowledgement of the local mums and some dads who form invisible networks of support the world over and keep it all going with love and laughs. This is for you (but you don t have to read it).

8

9 CONTENTS 1 Social Media Data Mining Becomes Ordinary 1 Data Abundance and Its Consequences 1 Researching Ordinary Social Media Data Mining 11 2 Why Study Social Media Data Mining? 19 Introduction 19 Four Characteristics of Social Media: Participation, Sharing, Intimacy and Monetisation 21 What Is Social Media Data Mining? 29 Conclusion 38 3 What Should Concern Us About Social Media Data Mining? Key Debates 41 Introduction 41 Critiques of (Social Media) Data Mining 44 Less Privacy, More Surveillance 44 Discrimination and Control 47 Methodological Concerns 49 Issues of Access and Inequality 52 Seeking Agency in Data Mining Structures 53 Worker Agency 56 User Agency 58 Techno-agency 60 ix

10 x CONTENTS Postscript on Agency: Acting Ethically in Times of Data Mining 62 Conclusion 64 4 Public Sector Experiments with Social Media Data Mining 67 Introduction 67 Action Research and the Production of Publics 70 Knowing and Forming Publics 70 Action-Researching Public Uses of Social Media Data Mining 71 Social Media Data Mining for the Public Good? 76 Uses of Social Media Data Mining 76 Understanding Publics, Desiring Numbers 79 Constituting Publics 87 How Keywords Constitute Publics 87 How Expertise (Or Its Absence) Constitutes Publics 89 Working Around Data Non-abundance to Constitute Publics 91 Conclusion: What Should Concern Us About Public Sector Social Media Data Mining? 94 5 Commercial Mediations of Social Media Data 99 Introduction 99 The Practices of Intermediary Insights Companies and the Concerns of Workers 104 Who Companies and Interviewees Are 104 What Companies and Interviewees Do 106 Social Media Data Mining as Moral and Economic Practice 111 Accessing Public Data 113 Drawing Ethical Boundaries 116 Transparency as Ethics 118 Who Benefits from Social Media Data Mining? 120 Regulation as Ethical Solution? 122 Conclusion: Concerns and Ethics in Commercial Social Media Data Mining Practice 124

11 CONTENTS xi 6 What Happens to Mined Social Media Data? 129 Introduction 129 Interviewees, Organisations, and Their Uses of Social Media Data Mining 132 The Consequences of Social Media Data Mining 137 Concrete Action and Organisational Complexity 137 Organisational Change and the Quality of Working Life 141 Data Evangelism and The Fetishism of the The Parasite on the Rhino? Reflections on Ethical Issues 151 Conclusion: Consequences and Concerns Fair Game? User Evaluations of Social Media Data Mining 159 Introduction 159 What Do Users Think? Studies of Users Views 162 Quantitative Studies of Attitudes to Digital Data Tracking 162 Qualitative Studies of Social Media User and Attitudes to Social Media Data Mining, and the Contextual Integrity Framework 166 Focus Group Methods for Researching User Perspectives 169 What Do Users Think? Focus Group Findings 171 Diverse Perspectives 171 Common Threads: Concerns About Fairness 179 What Concerns Users? Fairness, Transparency, Contextual Integrity Doing Good with Data: Alternative Practices, Elephants in Rooms 189 Introduction 189 Elephants in Rooms: Academic Social Media Data Mining 191 Overview of Academic Social Media Data Mining 191 Concerns and Criticisms 193 Un-Black-Boxing Social Media Data Mining 198

12 xii CONTENTS Alternative Practices: Data Activism 202 The Open Data Movement 202 Re-active and Pro-active Data Activism 205 Doing Good, Or Doing Bad, Through Data Activism? 212 Conclusion New Data Relations and the Desire for Numbers 221 Established Concerns 222 Emerging Concerns 224 The Desire for Numbers 224 (Not) Thinking Critically About Data-making 225 Work Effects 228 Doing Good with Data 230 New Data Relations 232 Doing Better with Data 234 Bibliography 237 Index 255

13 LIST OF FIGURES Figure 2.1 An illustration of mined social media data from the Museum Analytics website 30 Figure 2.2 The results of a search as displayed in NodeXL 34 Figure 2.3 Example of a Gephi visualisation 35 Figure 4.1 Networks of participants in relation to 1 month of activity data from the Facebook of one of the partners 80 Figure 4.2 Key contribution channels for one particular search, identified by one of the commercial tools 82 Figure 4.3 The results of a search as displayed in NodeXL 90 Figure 4.4 A sample interface from one of the commercial tools 92 Figure 6.1 Some of the content from social media insights reports circulated within interviewees organisations 145 Figure 6.2 Some of the content from social media insights reports circulated within interviewees organisations 146 Figure 6.3 Some of the content from social media insights reports circulated within interviewees organisations 147 Figure 6.4 Some of the content from social media insights reports circulated within interviewees organisations 147 Figure 6.5 Some of the content from social media insights reports circulated within interviewees organisations 148 xiii

14 Figure 6.6 Some of the content from social media insights reports circulated within interviewees organisations 148 Figure 8.1 A screenshot from Periscopic s animated visualisation of lives lost to gun-related deaths, ) 211 Figure 8.2 Findings from experiments into who tweets about open data 213 Figure 8.3 Findings from experiments into which terms are used alongside open data /#opendata 214

15 LIST OF TABLES Table 4.1 Social media platforms and data mining tools used by public sector organisations in which we carried out interviews 77 Table 5.1 Social media insights interviewee and company details (anonymised) 102 Table 6.1 Interviewees, their organisations, and social media insights tools used 134 Table 7.1 Focus groups and members 170 Table 7.2 Scenarios and participants initial responses 172 xv