Prospects for Detecting Deception on Twitter

This paper discusses the prospects for building a system that helps a human analyst to detect deception on Twitter. First, based on a previously published taxonomy of cyber deception, Twitter deception is examined. Second, a number of indicators are introduced. These are cues, helpful for a human analyst but not necessary or sufficient characteristics of deception. Third, with the indicators as a background, the prospects for deception detection are discussed. The paper represents a first step towards a systematic method for detecting deception on Twitter.