Too good to be true? Predicting author profiles from abusive language

The problem of online threats and abuse could potentially be mitigated with a computational approach, where sources of abuse are better understood or identified through author profiling. However, abusive language constitutes a specific domain of language for which it has not yet been tested whether differences emerge based on a text author's personality, age, or gender. This study examines statistical relationships between author demographics and abusive vs normal language, and performs prediction experiments for personality, age, and gender. Although some statistical relationships were established between author characteristics and language use, these patterns did not translate to high prediction performance. Personality traits were predicted within 15% of their actual value, age was predicted with an error margin of 10 years, and gender was classified correctly in 70% of the cases. These results are poor when compared to previous research on author profiling, therefore we urge caution in applying this within the context of abusive language and threat assessment.

[1]  M. Ashton,et al.  The HEXACO–60: A Short Measure of the Major Dimensions of Personality , 2009, Journal of personality assessment.

[2]  Dong Nguyen,et al.  "How Old Do You Think I Am?" A Study of Language and Age in Twitter , 2013, ICWSM.

[3]  Benno Stein,et al.  Overview of the 4th Author Profiling Task at PAN 2016: Cross-Genre Evaluations , 2016, CLEF.

[4]  Lin Qiu,et al.  You are what you tweet: Personality expression and perception on Twitter , 2012 .

[5]  J. Pennebaker,et al.  Linguistic styles: language use as an individual difference. , 1999, Journal of personality and social psychology.

[6]  P Kuppens,et al.  Categories versus dimensions in personality and psychopathology: a quantitative review of taxometric research , 2011, Psychological Medicine.

[7]  Gregory J. Park,et al.  Predicting Dark Triad Personality Traits from Twitter Usage and a Linguistic Analysis of Tweets , 2012, 2012 11th International Conference on Machine Learning and Applications.

[8]  T. Hastie,et al.  Generalized Additive Model Selection , 2015, 1506.03850.

[9]  Jon Oberlander,et al.  Whose Thumb Is It Anyway? Classifying Author Personality from Weblog Text , 2006, ACL.

[10]  Filiz Garip.  What failure to predict life outcomes can teach us , 2020, Proceedings of the National Academy of Sciences.

[11]  Danny Azucar,et al.  Predicting the Big 5 personality traits from digital footprints on social media: A meta-analysis , 2018 .

[12]  J. Pennebaker,et al.  PERSONALITY PROCESSES AND INDIVIDUAL DIFFERENCES Words of Wisdom: Language Use Over the Life Span , 2003 .

[13]  Pascale Fung,et al.  One-step and Two-step Classification for Abusive Language Detection on Twitter , 2017, ALW@ACL.

[14]  Carla J. Groom,et al.  Gender Differences in Language Use: An Analysis of 14,000 Text Samples , 2008 .

[15]  Lyle H. Ungar,et al.  Studying the Dark Triad of Personality through Twitter Behavior , 2016, CIKM.

[16]  Martine De Cock,et al.  A Multivariate Regression Approach to Personality Impression Recognition of Vloggers , 2014, WCPR '14.

[17]  Garth Davies,et al.  Searching for signs of extremism on the web: an introduction to Sentiment-based Identification of Radical Authors , 2018 .

[18]  Jordan B. Peterson,et al.  Personality and language use in self-narratives. , 2009 .

[19]  Ming-Wei Chang,et al.  BERT: Pre-training of Deep Bidirectional Transformers for Language Understanding , 2019, NAACL.

[20]  Ryan L. Boyd,et al.  The Development and Psychometric Properties of LIWC2015 , 2015 .

[21]  Yair Neuman,et al.  Profiling School Shooters: Automatic Text-Based Analysis , 2015, Front. Psychiatry.

[22]  Jennifer Golbeck,et al.  Predicting personality with social media , 2011, CHI Extended Abstracts.

[23]  Jeffrey Pennington,et al.  GloVe: Global Vectors for Word Representation , 2014, EMNLP.

[24]  Daniel N. Jones,et al.  Introducing the Short Dark Triad (SD3) , 2014, Assessment.

[25]  Isabelle van der Vegt,et al.  The temporal evolution of a far-right forum , 2020, J. Comput. Soc. Sci..

[26]  D. B. Skillicorn,et al.  A Bootstrapped Model to Detect Abuse and Intent in White Supremacist Corpora , 2020, 2020 IEEE International Conference on Intelligence and Security Informatics (ISI).

[27]  P. Read,et al.  Pseudocommando mass murderers: A big five personality profile using psycholinguistics , 2019, Current Psychology.

[28]  François Curtin,et al.  Multiple correlations and bonferroni’s correction , 1998, Biological Psychiatry.

[29]  Jacob Cohen.  Statistical Power Analysis for the Behavioral Sciences , 1969, The SAGE Encyclopedia of Research Design.

[30]  Lisa Kaati,et al.  Assessment of risk in written communication : Introducing the Profile Risk Assessment Tool (PRAT) , 2018 .

[31]  Lisa Kaati,et al.  Measuring online affects in a white supremacy forum , 2016, 2016 IEEE Conference on Intelligence and Security Informatics (ISI).

[32]  Dirk Hovy,et al.  Hateful Symbols or Hateful People? Predictive Features for Hate Speech Detection on Twitter , 2016, NAACL.

[33]  Joel R. Tetreault,et al.  Abusive Language Detection in Online User Content , 2016, WWW.