Modeling Longitudinal Behavior Dynamics Among Extremist Users in Twitter Data
Open Access
- Author:
- Murugan, Priyadarshini
- Graduate Program:
- Informatics
- Degree:
- Master of Science
- Document Type:
- Master Thesis
- Date of Defense:
- March 19, 2021
- Committee Members:
- Anna Cinzia Squicciarini, Thesis Advisor/Co-Advisor
Christopher H Griffin, Thesis Advisor/Co-Advisor
Dongwon Lee, Committee Member
Sarah Michele Rajtmajer, Committee Member
Mary Beth Rosson, Program Head/Chair - Keywords:
- Twitter
Extremist users
Chaos test
Hurst Exponent
Behavioral Patterns - Abstract:
- In this thesis, we use a dynamical systems perspective to analyze a collection of 2.4 million tweets known to originate from ISIS and ISIS-related users. From those users active over a long period of time (i.e., 2+ years), we derive sequences of behaviors and show that the top users cluster into 4 behavioral classes, which naturally describe roles within the ISIS communication structure. We then correlate these classes to the retweet network of the top users showing the relationship between dynamic behavior and retweet network centrality. We use the underlying model to formulate informed hypotheses about the role each user plays. Finally, we show that this model can be used to detect outliers, i.e., accounts that are thought to be outside the ISIS organization but seem to be playing a key communications role and have dynamic behavior consistent with ISIS members. This study further explores three datasets from a dynamical perspective as a preliminary analysis to differentiate human traffic and bot traffic. We perform a comparative study on three-time series datasets: ISIS dataset, Russian troll messages dataset, and Pennsylvania State Representative’s dataset. Through this analysis, we aim to identify if the ISIS users exhibit bot-like or human-like behavior based on the underlying dynamics using the 0-1 chaos test, Hurst Exponent, and Periodicity algorithm.