The Automated Detection of Trolling Bots and Cyborgs and the Analysis of Their Impact in the Social Media

Social media has become a place for discussion and debate on controversial topics, and thus provides an opportunity to influence public opinion. This possibility has given rise to a specific behavior known as trolling, which can be found in almost every discussion that includes emotionally appealing topics. A troll is an individual who shares inflammatory, extraneous or off-topic messages in social media, with the primary intent of provoking readers into an emotional response or otherwise disrupting on-topic discussion. Trolling is thus a useful tool for any organization willing to force a discussion off-track in the situations when one has no proper facts to back one's arguments. In this paper, the analysis of trolling is based on public discussion stakeholder classification by Luoma-Aho (2015), including positively engaged faith-holders, negatively engaged hateholders, and fakeholders. Trolls can be considered as either hateholders (humans) or fakeholders (bots or cyborgs). It is stated by Luoma-Aho that the influence of a fakeholder appears larger than it really is in practice, but tools for analyzing the impact are not provided in her work. This paper continues the work by Paavola and Jalonen (2015), who examined in their paper whether sentiment analysis could be utilized in detecting trolling behavior. It was concluded that sentiment analysis as such cannot detect trolls, but results indicated that social media analytics tools can generally be utilized for this task. In this paper the work continues with automatic detection of bots, which facilitates the analysis of fakeholder communication's impact. The automatic bot detection feature is implemented in the sentiment analysis tool in order to remove the noise in a discussion.