A text-mining tool to identify gossip motives

Gossip, information exchange about absent third parties, is ubiquitous in human societies. Recently, a conceptual framework to identify why people engage in this behavior was developed and supported empirically. This framework identifies five gossip motives: social enjoyment, information gathering and validation, negative influence, group protection, and emotion venting. Researchers, however, currently lack an automated tool that can allow them to easily/consistently identify gossip motives in unstructured texts. In addition, prior research has almost exclusively relied on self-reported gossip motives, which might provide an inaccurate reflection of intentions for behavior (due to social desirability bias or people’s inability to accurately judge their motives). Having a way to identify motives not relying on self-reported measures is an important step forward in the study of gossip across different fields. The text mining tool I plan to develop in collaboration with Terence Dores Cruz will address exactly these issues, while opening up a number of relevant research questions: a) what is the relationship between gossip motives and individual/group behaviors? b) How do gossip motives differ across situations? c) how independent are gossip motives from one another?


Data Science Project in FSW, IT Innovation Fund