Detecting Bots on Russian Political Twitter.

Area of Science:

• Computational Social Science
• Political Science
• Network Science

Background:

• Automated and semiautomated Twitter accounts (bots) are increasingly recognized for their influence on political discourse.
• Understanding bot activity is crucial for analyzing the integrity of online political discussions.

Purpose of the Study:

• To develop and apply a bot detection methodology to the Russian Twittersphere.
• To quantify bot activity during a critical period of Russian political events (February 2014 - December 2015).
• To identify characteristics distinguishing bots from human users and their primary activities.

Main Methods:

• Development of an ensemble of classifiers for Twitter bot detection.
• Application of the methodology to analyze tweets related to Russian politics.
• Statistical analysis to identify bot characteristics and predict bot activity.

Main Results:

• Bots produced over 50% of tweets on the majority of days concerning Russian politics.
• The software platform used for tweeting was a strong predictor of bot accounts.
• Bots were actively involved in disseminating news stories and promoting media outlets.

Conclusions:

• Bot activity significantly shapes political conversations on Twitter, particularly in the Russian context.
• The study provides a robust method for bot detection and characterization.
• Findings underscore the role of bots in information propagation and media promotion within political discussions.