Skip to main content
Peer-reviewed Authoritative

Overview of the CLPsych 2025 Shared Task: Capturing Mental Health Dynamics from Social Media Timelines

Overview of the CLPsych 2025 Shared Task, which combined longitudinal modeling of a person's mental-health state across their social-media timeline with evidence extraction and summarization. Its subtasks asked systems to extract text spans reflecting adaptive and maladaptive self-states, assign per-post well-being scores on a 1-10 scale, and summarize how self-states evolve at the post and timeline level.

Publisher

Association for Computational Linguistics (CLPsych 2025 Workshop)

Published

1 May 2025

Added

today

Key Findings

  • Pairs longitudinal mental-state tracking with extraction of the supporting evidence spans behind each self-state label
  • Scores well-being per post on a 1-10 scale and summarizes self-state dynamics at both the post and timeline level
  • Frames mental-health monitoring as evidence-grounded and human-interpretable rather than a single opaque score

Methodology Notes

Peer-reviewed shared-task overview, 10th CLPsych Workshop (NAACL 2025). Published May 2025; exact day not stated, so the day is set to 01. Substrate is social-media timelines, not chatbot conversation.

Sources

ACL Anthology (primary)

Archived snapshot (Wayback Machine) — preserved against link rot

Authors

Talia Tseriotou, Jenny Chim, Ayal Klein, Aviad Shamir, Guy Dvir, Iman Munire Bilal, George Kennedy, Chandan Kumar Singh Kohli, Anthony Hills, Ayah Zirikly, Dana Atzil-Slonim, Maria Liakata

Tags

clpsychmoments-of-changelongitudinalevidence-extractionliakata

Cite This

APA

Talia Tseriotou et al. (2025). Overview of the CLPsych 2025 Shared Task: Capturing Mental Health Dynamics from Social Media Timelines. Association for Computational Linguistics (CLPsych 2025 Workshop). https://aclanthology.org/2025.clpsych-1.16/