MindTailor: Personalized Emotional Support via Post History-Grounded Case Formulation and Collaborative Refinement

2026-06-20

AI Summary

This work addresses the limitation of existing emotional support approaches, which often neglect users’ historical experiences and thus struggle to deliver deeply empathetic and precisely tailored responses. To overcome this, the authors propose a novel paradigm that constructs case conceptualizations from users’ past posts and employs a multi-strategy counselor agent framework that iteratively refines supportive replies through collaborative optimization. This approach uniquely integrates individual history into the response generation process, leveraging large language models, multi-agent collaboration, and an LLM-as-a-Judge evaluation framework. The study introduces ReddiSupp, a new dataset comprising 798 Reddit user threads paired with their historical posts. Experimental results demonstrate that the proposed method significantly outperforms baseline systems across automatic metrics, expert evaluations, and real-user studies, achieving state-of-the-art performance in empathy, personalization, perceived understanding, and overall user preference.

Abstract

As mental health concerns continue to rise globally, social media has emerged as a vital space where individuals seek emotional support. While prior work on personalized emotional support has leveraged seekers' emotional states, personas, and situational context, these approaches primarily capture the seeker's current state, overlooking the formative experiences that shape present concerns. In this work, we propose MindTailor, a framework that generates personalized emotional support responses by constructing a case formulation from the seeker's post history and iteratively refining responses through collaborative critique among counselor agents grounded in distinct counseling strategies. To enable research on this history-aware task, we construct ReddiSupp, a dataset of 798 Reddit posts paired with seekers' prior post histories. Through LLM-as-a-Judge evaluation, expert human evaluation, and a user study with seekers, we demonstrate that MindTailor outperforms baselines across these evaluations, improving empathy, personalization, understanding, and achieving the highest overall preference.

Problem

• personalized emotional support
• post history
• case formulation
• mental health
• social media

Innovation

• case formulation
• post history
• collaborative refinement
• personalized emotional support
• counselor agents