To start off our SPIRIT Speaker Series, we are thrilled to invite Johannes Eichstaedt, PhD from Stanford University.

He will be presenting on “Digital Text and LLMs to Understand and Improve Individual and Community Mental Health”.

Dr. Eichstaedt is an Assistant Professor in Psychology and the Shriram Faculty Fellow at the Institute for Human-Centered Artificial Intelligence. He also directs the Computational Psychology and Well-Being lab and co-founded the World Well-Being Project at the University of Pennsylvania in 2011.

Talk Summary: The language shared online (e.g., social media) or in communication (e.g., SMS) reflects people’s thoughts and emotions. These data, processed through Natural Language Processing and machine learning, can predict mental health and psychological traits. For example, we have used Facebook statuses to forecast depression before clinical diagnosis in medical records. At the population level, analysis of billions of social media posts can identify health risk patterns. For example, geo-tagged Tweets can be used to estimate county heart disease mortality more accurately than traditional health risk factors and to monitor depression and anxiety trends across time and space. Moreover, we are exploring the use of Large Language Models (LLMs) to enhance mental health care, including delivering Cognitive Behavioral Therapy (CBT) and training therapists.