Language is a powerful means to communicate to the world and ourselves what we are thinking about, and why we behave in certain ways. Associating linguistic patterns with specific cognitive patterns and mental states can be an insightful tool in understanding the unique cognitive styles of individuals — not only can it help understand implicit behavioral traits such as a person’s judgment and decision-making abilities, reasoning skills, but it can also help model complex mental health conditions such as depression, PTSD, anxiety, mood disorders, etc. which can be used for diagnosis and therapy to improve our well-being. In this talk, I will introduce novel techniques to train precise models on social media discourse to capture individual cognitive styles. I will then present my work on language-based assessments for mental health, demonstrating an improved methodology for eliciting and capturing relevant discourse in mental health assessment.
Presented by: Vasudha Varadarajan, PhD student in the Computer Sciences Department
Special Collections Seminar Room, E-2340, is located on the second floor of the Melville Library.