Measuring Mental and Physical Health using Social Media
Tuesday, May 30
11 a.m.-12 p.m.
Malone 228, Johns Hopkins University Homewood Campus
Zoom: https://wse.zoom.us/j/93478802403

Abstract: The content shared on social media is among the largest data sets on human behavior in history. We leverage this data stream to unobtrusively measure the mental and physical health and well-being of individuals and populations using machine learning and Natural Language Processing (NLP). For depression, machine learning models applied to patients’ Facebook statuses can predict their future depression status before it appears in their medical records. For heart disease, the leading cause of death, prediction models derived from geo-tagged Tweets can estimate county mortality rates better than traditional health risk factors. For well-being, Twitter can be used to estimate the subjective well-being of nations while controlling for sampling biases, with resolution unmatched by any other method. In emerging work, Johannes explores the use of Large Language Models to promote well-being and mental health, and highlights some of the many psychological and social impacts LLMs will have on society.

Bio: Johannes Eichstaedt is an Assistant Professor (Research) in Psychology at Stanford and a Shriram Faculty Fellow at the Institute for Human-Centered A.I., where he directs the Computational Psychology & Well-Being Lab. He is a computational psychologist and interdisciplinary data scientist studying the mechanisms that give rise to mental and physical health by applying NLP to digital text. In 2011, he co-founded the World Well-Being Project; with this inter-institutional consortium, he has since attracted $4.9m+ in funding and published 60+ articles. Johannes received his Ph.D. at the University of Pennsylvania in 2017, and in 2022 was elected a Rising Star by the Association of Psychological Science.