ANALYZING SOCIAL MEDIA NETWORKS WITH NODEXL INSIGHTS FROM A CONNECTED WORLD ANALYZING SOCIAL MEDIA NETWORKS WITH NODEXL INSIGHTS FROM A CONNECTED WORLD Derek L. Hansen Ben Shneiderman Marc A. Smith AmsterdAm • Boston • HeidelBerg • london new York • oxford • PAris • sAn diego sAn frAncisco • singAPore • sYdneY • tokYo morgan kaufmann is an imprint of elsevier Acquiring Editor: Mary James Development Editor: David Bevans Project Manager: Julie Ochs Designer: Dennis Schaefer Morgan Kaufmann is an imprint of Elsevier 30 Corporate Drive, Suite 400, Burlington, MA 01803, USA Copyright © 2011 Elsevier Inc. All rights reserved No part of this publication may be reproduced or transmitted in any form or by any means, electronic or mechanical, including photocopying, recording, or any information storage and retrieval system, without permission in writing from the publisher. Details on how to seek permission, further information about the Publisher’s permissions policies and our arrangements with organizations such as the Copyright Clearance Center and the Copyright Licensing Agency, can be found at our website: www.elsevier.com/permissions. This book and the individual contributions contained in it are protected under copyright by the Publisher (other than as may be noted herein). Notices Knowledge and best practice in this field are constantly changing. As new research and experience broaden our understanding, changes in research methods or professional practices, may become necessary. Practitioners and researchers must always rely on their own experience and knowledge in evaluating and using any information or methods described herein. In using such information or methods they should be mindful of their own safety and the safety of others, including parties for whom they have a professional responsibility. To the fullest extent of the law, neither the Publisher nor the authors, contributors, or editors assume any liability for any injury and/or damage to persons or property as a matter of products liability, negligence or otherwise, or from any use or operation of any methods, products, instructions, or ideas contained in the material herein. Library of Congress Cataloging-in-Publication Data Hansen, Derek L. Analyzing social media networks with NodeXL : insights from a connected world/Derek L. Hansen, Ben Shneiderman, Marc A. Smith. p. cm. Includes bibliographical references and index. ISBN 978-0-12-382229-1 (pbk. : alk. paper) 1. Data mining—Computer programs. 2. Information visualization-Computer programs. 3. Webometrics—Computer programs. 4. Online social networks. 5. NodeXL. I. Shneiderman, Ben. II. Smith, Marc A., 1965– III. Title. QA76.9.D343.H45 2010 006.7954—dc22 2010019806 British Library Cataloguing-in-Publication Data A catalogue record for this book is available from the British Library. ISBN: 978-0-12-382229-1 Printed in China 10 11 12 13 14 10 9 8 7 6 5 4 3 2 1 For information on all MK publications visit our website at www.mkp.com Derek: To Maren Hansen Ben: To Jennifer Preece Marc: To my strong ties to Madeline, Eli, and Christine Preface We live in the era of networks. Occasionally the aspi- could substantially lower the barrier to entry for social rations of academic researchers are in harmony with media network analysis while at the same time raising the needs of commercial developers, business entrepre- the power offered to users seeking network insights. We neurs, and government agency staffers. In our case, the chose Excel 2007 as our first platform for data import authors brought complementary backgrounds in infor- and manipulation because of its widespread acceptance mation studies, computer science, and sociology, as well in the business community, rich set of features, and con- as a shared interest in the interdisciplinary topics of venience in extending it. Then we took the best features human-computer interaction, network analysis, social from existing analysis and graphics tools, closely inte- media, and information visualization. grating statistical methods with information visualiza- At the same time, these topics were gaining interest tion strategies while keeping in mind the goal of being from the growing community of entrepreneurs who easy to learn by a wide range of users. were coping with the success of their social media com- The need for precise management of the visual prop- mercial platforms such as Twitter, Facebook, Flickr, erties of a network chart led to rich controls for vertex YouTube, and the equally remarkable open source layout, size, color, shape, images, labels, and opacity, communities that produced valuable resources such as and parallel attributes such as thickness for edges. The Wikipedia. These entrepreneurs perceived of their cre- desire for rapid filtering was satisfied with dynamic ations as networks, drawing on the language of social queries based on the ambitious use of double-slider communities and asking questions about motivation, range controls. The aspiration to support multiple influence, and social roles. sources of data led to a strong effort to import mul- The tools for social media network analysis and tiple networks from common platforms. The desire to visualization have been emerging from many research gain insights into social media networks led to integra- groups and startup companies. These pioneering net- tion of several “spigots” that open streams of network work analysis tools often require programming skills and data from popular social media network sources like knowledge of technical network terminology, making it a YouTube, Twitter, Flickr, and personal email. challenge for those without programming skills to import As NodeXL emerged in release after release, some- and make sense of network data. By 2000, network met- times only weeks apart, we tested it on our own social rics had become a mature topic, but research on layout network research projects and then used it in three and clustering algorithms has expanded dramatically in academic courses, plus professional tutorials at confer- the past decade, producing breakthroughs that raised ences. These experiences led us to improve usability the quality of what was possible to visualize. Similarly, aspects and also led to important clarifications of the strategies for filtering, highlighting, and de-cluttering analysis process that improved our teaching strategies. networks have matured as more users tackle a broader In each stage, we raised our aspirations about what variety of problems with increasingly large networks. we and our users could accomplish in terms of gain- The authors have been fortunate to be part of a ing actionable insights from social media analytics. team with unique skills that developed the NodeXL Although we are pleased with what NodeXL users are tool. Funded by Microsoft Research External Research able to accomplish, we are humbled by the richness and Projects group, the team is distributed across the United diversity of social media analytical possibilities. The States and Europe with links to others around the opportunities and challenges are substantial, which will world. Our weekly phone calls over more than 2 years keep researchers and developers productively engaged developed requirements and set priorities so that we for many years to come. ix Acknowledgments The authors would like to thank the many people the tool. We appreciate the time and attention our users who have made this document and the NodeXL proj- give the tool and the project and hope they will con- ect possible. First, the members of the NodeXL design tinue to upgrade with us as the project grows. We are and development team include Natasa Milic-Frayling, grateful to these and many other people for their efforts Eduarda Mendes Rodrigues, Janez Brank, and Annika to make NodeXL an easy and useful tool for under- Hupfeld from Microsoft Research in Cambridge, standing complex networks. England, and Jure Leskovec at Standford. We thank Tony Capone from Microsoft Research in Redmond, Andrew Fiore, the University of California at Berkeley Washington, for his remarkable programming prowess, Barry Wellman, University of Toronto thoughtful discussions about features, and always cour- Bill Johnston, ForumOne Communications teous help to us and the NodeXL user community. Cliff Lampe, Michigan State University Support for NodeXL development has been gener- Dave Ashton, OsVisNet Project ously provided by Tony Hey, Daron Green, and Dan David Shamma, Yahoo! Research Fay from the Microsoft Research External Research Eleanor Wynn, Intel Corporation Programs group in Redmond, Washington. Elizabeth Bonsignore, University of Maryland We thank Eric Gleave from the University of George Barnett, University of California at Davis Washington; Vladimir Barash from Cornell University; Han Woo Park, Yeungnam University and Cody Dunne, Udayan Khurana, and Adam Perer Howard Rheingold, social media author and teacher from the University of Maryland for their intellectual Jes Koepfler, University of Maryland contributions to our grand adventure. John Kelly, Morningside Analytics We thank SeriousEats (www.seriouseats.com), which Julian Hopkins, Monash University has allowed us to use data collected from its fascinating Keith Hampton, University of Pennsylvania online community. We also thank Emily Mason, Chad Lada Adamic, University of Michigan Doran, and Rachel Collins, who collected data sets Libby Hemphill, Illinois Institute of Technology used in the book as part of their coursework and came Mathieu Bastian, Project Gephi up with compelling analyses of them. Special thanks to Martha Russell, Stanford Media-X program Chris Wilson of Slate Magazine for sharing the Senate Nicky Van Zyl, Vodacom South Africa 2007 voting data. Nosh Contractor, Northwestern University Our many users have provided remarkable feedback, Phillip Howard, University of Washington but Pierre de Vries merits a specific mention for push- Pierre de Vries, independent telecommunications policy ing NodeXL beyond our expectations. Our research col- researcher laborators Dana Rotman and Elizabeth Bonsignore have Randy Farmer, social media strategy consulting, made it possible to field-test NodeXL and carefully doc- WikiAnswers ument the results. The students of several classes who Rob Cross, University of Virginia were assigned projects with NodeXL have been patient Ton Zijlstra and forgiving as we refined the rough edges. Many Yann Leroux thousands of people have downloaded NodeXL, and several have created research and business results using Our Twitter fans! xi About the Authors Derek L. Hansen is an assistant professor at the Visualization (2003). His book Leonardo’s Laptop: Human University of Maryland’s iSchool and director for the Values and the New Computing Technologies appeared in Center for the Advanced Study of Communities and October 2002 (MIT Press) (http://mitpress.mit.edu/ Information (http://casci.umd.edu), a multidisciplinary leonardoslaptop) and won the IEEE book award for research center focused on harnessing the power of novel Distinguished Literary Contribution. social technologies to support the needs of real and virtual Marc A. Smith is a sociologist specializing in the communities. He is also an active member of the Human social organization of online communities and comput- Computer Interaction Lab (www.cs.umd.edu/hcil). er-mediated interaction. He founded and managed the Dr. Hansen completed his Ph.D. from the University Community Technologies Group at Microsoft Research of Michigan’s School of Information where he was an in Redmond, Washington, and led the development of National Science Foundation–funded interdisciplinary social media reporting and analysis tools for Telligent STIET Fellow (http://stiet.si.umich.edu) focused on Systems. Smith leads the Connected Action consulting understanding and designing effective online sociotech- group and lives and works in Silicon Valley, California. nical systems. His research and teaching focus on mass Smith is the co-editor, with Peter Kollock, of collaboration, information reuse, consumer health infor- Communities in Cyberspace (Routledge), a collection of matics, and social network analysis of online communi- essays exploring the ways identity, interaction, and ties. One line of research focuses on helping community social order develop in online groups. analysts make sense of the mass of social data available Smith’s research focuses on computer-mediated col- through social media sites. Another line of research lective action: the ways group dynamics change when applies those methods to improve our understanding of they take place in and through social cyberspaces. Many best practices for supporting mass collaboration in the “groups” in cyberspace produce public goods and orga- medical, scientific, and entertainment domains. Finally, nize themselves in the form of a commons (for related Dr. Hansen is involved with designing novel tools that papers see http://delicious.com/marc_smith/Paper). take advantage of the unique properties of information Smith’s goal is to visualize these social cyberspaces, technologies. For example, he has done work on “col- mapping and measuring their structure, dynamics, and lection recommender systems” and social networking life cycles. At Microsoft, he developed the “Netscan” applications designed to anonymously disseminate web application and data mining engine that allows information about stigmatized illnesses through social researchers studying Usenet newsgroups and related networking sites. repositories of threaded conversations to get reports Ben Shneiderman (www.cs.umd.edu/~ben) is on the rates of posting, posters, crossposting, thread a professor in the Department of Computer Science length, and frequency distributions of activity. Smith and founding director (1983–2000) of the Human- applied this work to the development of a generalized Computer Interaction Laboratory (www.cs.umd.edu/ community analysis platform for Telligent, providing a hcil) at the University of Maryland. He was elected as web-based system for groups of all sizes to discuss and a Fellow of the Association for Computing (ACM) in publish their material to the web and analyze the emer- 1997 and a Fellow of the American Association for the gent trends that result. He contributes to the open and Advancement of Science (AAAS) in 2001. He received free NodeXL project (www.codeplex.com/nodexl) that the ACM SIGCHI Lifetime Achievement Award in 2001. adds social network analysis features to the familiar Shneiderman is the co-author, with Catherine Excel spreadsheet. A tutorial on social network analysis Plaisant, of Designing the User Interface: Strategies is evolving into a book and is freely available (http:// for Effective Human-Computer Interaction (5th ed., casci.umd.edu/NodeXL_Teaching). NodeXL enables April 2009, www.awl.com/DTUI). With S. Card and social network analysis of email, Twitter, Flickr, and J. Mackinlay, he co-authored “Readings in Information other network data sets. Visualization: Using Vision to Think” (1999). With The Connected Action consulting group (www.con- Ben Bederson he co-authored The Craft of Information nectedaction.net) applies social science methods in xiii xiv About the Authors general and social network analysis techniques in par- Smith received a B.S. in International Area Studies ticular to enterprise and Internet social media usage. from Drexel University in Philadelphia in 1988, an SNA analysis of data from message boards, blogs, wikis, M. Phil. in social theory from Cambridge University friend networks, and shared file systems can reveal in 1990, and a Ph.D. in Sociology from UCLA in 2001. insights into organizations and processes. Community He is an affiliate faculty at the Department of Sociology managers can gain actionable insights into the vol- at the University of Washington and the College of umes of community content created in their social Information Studies at the University of Maryland. media repositories. Mobile social software applications can visualize patterns of association that are otherwise invisible. Contributors Robert Ackland Australian Demographic and Social Bernie Hogan Oxford Internet Institute, University of Oxford, Research Institute, Australian National University, Oxford, UK Canberra, Australia Eduarda Mendes Rorigues Department of Informatics Vladimir Barash Information Science, Cornell University, Engineering, University of Porto, Porto, Portugal Ithaca, NY Natasa Milic-Frayling Microsoft Research Ltd., Cambridge, Laura W. Black School of Communications Studies, Ohio UK University, Athens, OH Dana Rotman University of Maryland, College Park, MD Dan Cosley Information Science, Cornell, University, Patrick Underwood Department of Sociology, University of Ithaca, NY Washington, Seattle, WA Jennifer Golbeck University of Maryland, College Park, MD Howard T. Welser Department of Sociology and Scott Golder Department of Sociology, Cornell University, Anthropology, Ohio University, Athens, OH Ithaca, NY xv