Pilot Evaluation of Crisis Support over Messenger (CSOM)

Dr. Gallo will serve as the subcontract PI at Northwestern University. He will develop and test a machine learning method that analyzes text conversations from Crisis Support Facebook Messenger (CSOM). The goal of this computer-based method is to automatically classify CSOM conversations into two categories: those conversations that reduced suicide risk and those that did not. Dr. Gallo and the Northwestern team will be responsible for evaluating the accuracy of the machine learning approach to classify CSOM conversation using several linguistic features including emotional content, language models, and part of speech tags. Dr. Gallo and the Northwestern team will participate in the production of publications that describes this methodology and present future results.