Hybrid elicitation of latent intent in open societies (HELIOS)

The reach of ubiquitous social networking tools enabled through new technology is evident in the recent regime change and the increased unrest within the Middle East. Such rapid dissemination of information may engender an equally rapid emergence of virtual communities of interest - some with potentially hostile intent. Twitter's accessibility and popularity have attracted a large number of automated programs, known as bots, which can serve as a double-edged sword within Twitter. While legitimate bots generate a large amount of benign tweets delivering news and updating information, malicious bots spread misinformation with potentially virulent results. This paper blends methods from traditional statistical classification methods with latent Dirichlet allocation with techniques of social network analysis to underpin threat assessment based upon similarities discerned among topic sets that have distilled by semantic analysis of Twitter intercepts.