DOI: 10.1093/schbul/sbag159.053 ISSN: 0586-7614

Building a Mental Health Support System for Rural Communities: An Intelligent Framework for Detecting Anxiety, Depression, and Psychological Stress via Dialect Semantic Understanding

Feng Li

Abstract

Objective

This study aims to develop and validate an intelligent system for recognizing key mental health states—specifically anxiety, depression, and psychological stress—among rural residents by leveraging dialect semantic analysis. It seeks to construct a targeted service delivery mechanism to provide timely emotional management and mental well-being support in underserved areas.

Subjects and Methods

The study involved 500 participants across three dialect-diverse rural regions. A multimodal dataset included audio recordings of daily conversations, mental health scales (PHQ-9, GAD-7, PSS) screening for emotional disorders, anxiety, and stress, and digital logs. A deep learning pipeline was developed for dialect-adaptive speech recognition, semantic-emotion mapping, and context-aware service recommendation, tested in a 6-month field trial.

Results

The dialect-aware recognition model achieved 82.4% accuracy in identifying distress states associated with anxiety and depression, outperforming a standard model (68.1%). The system successfully delivered personalized interventions—such as counseling referrals and stress management resources—with 75% user acceptance. Post-trial data showed a 30% reduction in reported isolation and a 25% increase in mental health service utilization.

Conclusions

An intelligent system based on dialect understanding can effectively identify early signs of anxiety, depression, and stress, enabling proactive mental health support in rural communities. This approach bridges linguistic and cultural gaps in digital mental health care, offering a scalable model for emotional management and psychological well-being promotion.

Fund Project

The final outcome of the Key Project of Liaoning Provincial Social Science Planning Fund in 2023 (Project Number: L23AYY006).

Corresponding Author

Feng Li, College of International Cultural Exchange, Dalian University, Dalian 116622, China.

More from our Archive