Taxonomy and Bot Threats in Social Networks

As the impact of social services is continuously growing, there increase in various security threats that can use social bots. At the same time, existing social threat taxonomies leverage suspicious actions of account or behavioral patterns. As the result, known threat classifications are almost always unrelated to bot classifications. In this paper, we propose a taxonomy of bots, which differs from the existing ones as it considers not only the functional bot features but also scenarios for bot creation and management. We also propose the classification of bot threats, which was validated with a survey method. For that, we made a survey with the participation of experts who study social bots. Based on their assessments, we created an integral score for each type of threat. The presented taxonomy can be used in bot detection models and risk analysis to assess the risk of threats.