Abstract

Suicidal rates have been increasing since 2000 according the latest report of Centers for disease control and prevention. Nowadays, when internet opens a channel where people can communicate and information remains registered, acoustic, semantic and syntactic analyses are being exhaustively explored to try to find hidden cues that can be used to detect signs of different mental conditions. In this work, to develop a method to detect suicidal signs, we analyze poems from poets who committed suicide. We use bipartite graph matching algorithms after data retrieval to assure our results are not susceptible to bias, and focus on linguistic content (e.g. similarity to specific concepts) and structure (e.g. density of ideas). Our classification results using different classifiers yield accuracy rates of up to 86% to discriminate suicidal from control poets.

Team

Carla Agurto, Computational Biology Center, T.J. Watson IBM Research Laboratory
Pat Pataranutaporn, Arizona State University
Elif K. Eyigoz, Computational Biology Center, T.J. Watson IBM Research Laboratory
Gustavo Stolovitzky, Computational Biology Center, T.J. Watson IBM Research Laboratory.
Guillermo Cecchi, Computational Biology Center, T.J. Watson IBM Research Laboratory

Publication

Agurto, C., Pataranutaporn, P., Eyigoz, E. K., Stolovitzky, G., & Cecchi, G. (2018). Predictive linguistic markers of suicidality in poets. In IEEE International Conference on Semantic Computing.