Detecting Potential Cyber Armies of Election Campaigns Based on Behavioral Analysis

摘要

Recently, online social networks have been popular for election campaigns to monitor public opinion, spread information, and even try to influence the discussion among the platforms. In this study, we focus on the collusion of potential political teams or individuals who attempt to produce discussions and a series of positive/negative comments to support/attack specific candidates. We collect the user behavior in the most extensive online forum in Taiwan before a national election and use statistical analysis to identify such users. We also verify the results by manually reading the published content of the users. From the results, we hope this study can benefit users to identify underneath information manipulation and retain the trustiness of online society.