Bots (i.e. automated accounts) involve in social campaigns typically for two obvious reasons: to inorganically sway public opinion and to build social capital exploiting the organic popularity of social campaigns. In this work, we use DeBot to detect a large number of bots interested in politics, sports and e-commerce. We perform multiaspect (i.e. temporal, textual, and topographical) clustering of bots, and ensemble the clusters to identify campaigns of bots. We observe similarity among the bots in a campaign in various aspects such as temporal correlation, sentimental alignment, and topical grouping. However, we also discover bots compete in gaining attention from humans, follow leads from human users, and occasionally engage in arguments. We classify such bot interactions in two primary groups: agreeing (i.e. positive) and disagreeing (i.e. negative) interactions. We develop an automatic interaction classifier to discover novel interactions among bots participating in social campaigns. We apply the classifier successfully on three domains: politics, sports, and e-commerce.
The code is available on my Github.
please email nabuelrub AT unm DOT edu to request access to more data \or if you have any questions .
Pdf version of the paper is available here