Mental Health Impacts of AI Companions
AI-powered companions are increasingly used as tools for emotional support, yet their effects on long-term mental health remain poorly understood. This project triangulates quasi-experimental evidence from social media, user perspectives, and relational theory to examine whether AI companionship alleviates or exacerbates loneliness and isolation.
ACM CHI 2026 — Yuan, Y., Zhang, J., Aledavood, T., Zhang, R., & Saha, K. Mental Health Impacts of AI Companions: Triangulating Social Media Quasi-Experiments, User Perspectives, and Relational Theory.
In Proceedings of the CHI Conference on Human Factors in Computing Systems. [
Link]
Mental Health Coping Narratives on Social Media
Online communities host large-scale discussions of mental health coping strategies, offering a window into population-level behavior. This project investigates whether exposure to first-person recovery and coping stories produces a Papageno effect — reducing suicidal ideation and self-harm — using causal inference methods on social media data.
WWW 2023 — Yuan, Y., Saha, K., Keller, B., Isometsä, E. T., & Aledavood, T. (2023). Mental Health Coping Stories on Social Media: A Causal-Inference Study of Papageno Effect.
In Proceedings of the ACM Web Conference 2023 (pp. 2677–2685). [
Link]
IC2S2 2023 — Yuan, Y., Saha, K., Keller, B., Isometsä, E. T., & Aledavood, T. Does Papageno Effect Occur on Social Media?. 9th International Conference on Computational Social Science.
Marginalized Communities and Online Well-being
Minority stress — the chronic stress faced by stigmatized social groups — manifests in online spaces in measurable ways. This project uses large-scale social media analysis to study how marginalized communities, including LGBTQ individuals and people struggling with substance use, express and cope with stress, with the aim of informing targeted mental health interventions.
ICWSM 2023 — Yuan, Y., Verma, G., Keller, B., & Aledavood, T. (2023). Minority Stress Experienced by LGBTQ Online Communities during the COVID-19 Pandemic.
Proceedings of the International AAAI Conference on Web and Social Media (Vol. 17, pp. 936–947). [
Link]
JMIR Formative Research 2024 — Yuan, Y., Kasson, E., Taylor, J., Cavazos-Rehg, P., De Choudhury, M., & Aledavood, T. (2024). Examining the Gateway Hypothesis and Mapping Substance Use Pathways on Social Media: A Machine Learning Approach.
JMIR Formative Research, 8, e54433. [
Link]