Topic Modeling Reveals Distinct Interests within an Online Conspiracy Forum

Conspiracy theories play a troubling role in political discourse. Online forums provide a valuable window into everyday conspiracy theorizing, and can give a clue to the motivations and interests of those who post in such forums. Yet this online activity can be difficult to quantify and study. We describe a unique approach to studying online conspiracy theorists which used non-negative matrix factorization to create a topic model of authors' contributions to the main conspiracy forum on Reddit.com. This subreddit provides a large corpus of comments which spans many years and numerous authors. We show that within the forum, there are multiple sub-populations distinguishable by their loadings on different topics in the model. Further, we argue, these differences are interpretable as differences in background beliefs and motivations. The diversity of the distinct subgroups places constraints on theories of what generates conspiracy theorizing. We argue that traditional “monological” believers are only the tip of an iceberg of commenters. Neither simple irrationality nor common preoccupations can account for the observed diversity. Instead, we suggest, those who endorse conspiracies seem to be primarily brought together by epistemological concerns, and that these central concerns link an otherwise heterogenous group of individuals.

[1]  H. Sebastian Seung,et al.  Learning the parts of objects by non-negative matrix factorization , 1999, Nature.

[2]  Andrew Parker,et al.  Conspiracy theory and cognitive style: a worldview , 2015, Front. Psychol..

[3]  Martin Voracek,et al.  Conspiracist ideation in Britain and Austria: evidence of a monological belief system and associations between individual psychological differences and real-world and fictitious conspiracy theories. , 2011, British journal of psychology.

[4]  Patrick J. Leman,et al.  The born conspiracy , 2007 .

[5]  Mark Dredze,et al.  Discovering Shifts to Suicidal Ideation from Mental Health Content in Social Media , 2016, CHI.

[6]  Adrian Vermeule,et al.  Conspiracy Theories , 2008 .

[7]  Stefan A. Ortlieb,et al.  Thirty shades of truth: conspiracy theories as stories of individuation, not of pathological delusion , 2013, Front. Psychol..

[8]  Sander van der Linden,et al.  The conspiracy-effect: Exposure to conspiracy theories (about global warming) decreases pro-social behavior and science acceptance , 2015 .

[9]  Stephan Lewandowsky,et al.  NASA Faked the Moon Landing—Therefore, (Climate) Science Is a Hoax , 2013, Psychological science.

[10]  C. Sunstein,et al.  Conspiracy Theories: Causes and Cures* , 2009 .

[11]  Xujuan Zhou,et al.  Using social connection information to improve opinion mining: Identifying negative sentiment about HPV vaccines on Twitter , 2015, MedInfo.

[12]  Isabelle Guyon,et al.  Clustering: Science or Art? , 2009, ICML Unsupervised and Transfer Learning.

[13]  A. Furnham,et al.  Analytic thinking reduces belief in conspiracy theories , 2014, Cognition.

[14]  Gaël Varoquaux,et al.  Scikit-learn: Machine Learning in Python , 2011, J. Mach. Learn. Res..

[15]  Ewan Klein,et al.  Natural Language Processing with Python , 2009 .

[16]  Guido Caldarelli,et al.  Trend of Narratives in the Age of Misinformation , 2015, PloS one.

[17]  Robert Tibshirani,et al.  Estimating the number of clusters in a data set via the gap statistic , 2000 .

[18]  Helen A. Berger,et al.  Persuasions of the Witch's Craft: Ritual Magic in Contemporary England. , 1990 .

[19]  K. Mandl,et al.  Associations Between Exposure to and Expression of Negative Opinions About Human Papillomavirus Vaccines on Social Media: An Observational Study , 2015, Journal of medical Internet research.

[20]  Ted Goertzel,et al.  Belief in Conspiracy Theories , 1994 .

[21]  Robert S. Singh The Farrakhan Phenomenon: Race, Reaction, and the Paranoid Style in American Politics , 1997 .

[22]  George A. Miller,et al.  WordNet: A Lexical Database for English , 1995, HLT.

[23]  A. Cuevas,et al.  Estimating the number of clusters , 2000 .

[24]  Karen M. Douglas,et al.  Dead and Alive , 2012, Diary of a Detour.

[25]  Christos Boutsidis,et al.  SVD based initialization: A head start for nonnegative matrix factorization , 2008, Pattern Recognit..

[26]  D. Kahan Climate-Science Communication and the Measurement Problem , 2015 .

[27]  G. Caldarelli,et al.  The spreading of misinformation online , 2016, Proceedings of the National Academy of Sciences.

[28]  David Buttler,et al.  Exploring Topic Coherence over Many Models and Many Topics , 2012, EMNLP.

[29]  G. Menzies,et al.  1421: The Year China Discovered The World , 2002 .

[30]  Kam-Fai Wong,et al.  Interpreting TF-IDF term weights as making relevance decisions , 2008, TOIS.

[31]  T. V. Pollet,et al.  Political Extremism Predicts Belief in Conspiracy Theories , 2015 .

[32]  P. Rousseeuw Silhouettes: a graphical aid to the interpretation and validation of cluster analysis , 1987 .

[33]  Adam D. Galinsky,et al.  Lacking Control Increases Illusory Pattern Perception , 2008, Science.

[34]  David S. G. Goodman Mao and The Da Vinci Code: conspiracy, narrative and history , 2006 .

[35]  Karen M. Douglas,et al.  Online communication as a window to conspiracist worldviews , 2015, Front. Psychol..