What is Stigma Attributed to? A Theory-Grounded, Expert-Annotated Interview Corpus for Demystifying Mental-Health Stigma

2025-05-19

AI Summary

Mental health stigma impedes help-seeking behavior, yet existing computational studies are hindered by a lack of theoretically grounded, culturally sensitive annotated corpora. To address this, we introduce the first theory-driven, expert-annotated human–machine interview corpus—comprising 4,141 dialogues from 684 participants with diverse sociocultural backgrounds—designed using Attribution Theory to support fine-grained labeling of stigma-related attributions. Methodologically, we integrate social-psychological frameworks with authentic interview data, moving beyond reliance on social media or synthetic texts; we employ a multi-tier expert annotation protocol and benchmark performance using BERT, RoBERTa, and related models. Experiments reveal significant model biases in distinguishing culturally nuanced attributions (e.g., “laziness” vs. “pathological condition”). The corpus is publicly available via the ACL Anthology and has been adopted in multiple downstream studies.

Technology Category

• Application Domains: Humanities & Computational Social Science
• Natural Language Processing: Sentiment Analysis, Stylistic Analysis, and Argument Mining
• Cognitive Modeling & Cognitive Systems: Social Cognition And Interaction

Application Category

• Economics, Online Markets and Human Computation: Data quality aspects of human-annotated datasets
• Search and Retrieval-Augmented AI: Multilingual and cross-lingual Web search
• Semantics and Knowledge: Data modeling to support human-machine intelligence, including LLMs agents, intelligent system behavior, explanations, and user-friendly interactions

Abstract

Mental-health stigma remains a pervasive social problem that hampers treatment-seeking and recovery. Existing resources for training neural models to finely classify such stigma are limited, relying primarily on social-media or synthetic data without theoretical underpinnings. To remedy this gap, we present an expert-annotated, theory-informed corpus of human-chatbot interviews, comprising 4,141 snippets from 684 participants with documented socio-cultural backgrounds. Our experiments benchmark state-of-the-art neural models and empirically unpack the challenges of stigma detection. This dataset can facilitate research on computationally detecting, neutralizing, and counteracting mental-health stigma.

Problem

• Understanding causes of mental-health stigma
• Lack of theory-based stigma classification datasets
• Detecting and mitigating stigma computationally

Innovation

• Expert-annotated theory-informed interview corpus
• State-of-the-art neural models benchmarking
• Computational detection and neutralization of stigma