Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

Psy-LLM: Scaling up Global Mental Health Psychological Services with AI-based Large Language Models

Tin Lai, Yukun Shi, Zicong Du, Jiajie Wu, Ken Fu, Yichao Dou, Ziqi Wang

TL;DR

This paper introduces Psy-LLM, an AI-assisted framework for online psychological consultation designed to alleviate the demand on mental health professionals. It combines PanGu-350M and WenZhong-based large-language models with a curated PsyQA dataset and crawled psychology content, then evaluates performance with both automated metrics and human judgments. A web-based frontend with secure, modular architecture demonstrates the system's practicality for frontline support and triage, while the discussion candidly addresses data quality, ethical considerations, and real-world deployment challenges. Results show PanGu-based outputs outperform WenZhong across multiple metrics, though still fall short of human-level performance, underscoring the need for higher-quality data, improved architectures, and careful integration with professional care.

Abstract

The demand for psychological counselling has grown significantly in recent years, particularly with the global outbreak of COVID-19, which has heightened the need for timely and professional mental health support. Online psychological counselling has emerged as the predominant mode of providing services in response to this demand. In this study, we propose the Psy-LLM framework, an AI-based assistive tool leveraging Large Language Models (LLMs) for question-answering in psychological consultation settings to ease the demand for mental health professions. Our framework combines pre-trained LLMs with real-world professional Q\&A from psychologists and extensively crawled psychological articles. The Psy-LLM framework serves as a front-end tool for healthcare professionals, allowing them to provide immediate responses and mindfulness activities to alleviate patient stress. Additionally, it functions as a screening tool to identify urgent cases requiring further assistance. We evaluated the framework using intrinsic metrics, such as perplexity, and extrinsic evaluation metrics, with human participant assessments of response helpfulness, fluency, relevance, and logic. The results demonstrate the effectiveness of the Psy-LLM framework in generating coherent and relevant answers to psychological questions. This article discusses the potential and limitations of using large language models to enhance mental health support through AI technologies.

Psy-LLM: Scaling up Global Mental Health Psychological Services with AI-based Large Language Models

TL;DR

This paper introduces Psy-LLM, an AI-assisted framework for online psychological consultation designed to alleviate the demand on mental health professionals. It combines PanGu-350M and WenZhong-based large-language models with a curated PsyQA dataset and crawled psychology content, then evaluates performance with both automated metrics and human judgments. A web-based frontend with secure, modular architecture demonstrates the system's practicality for frontline support and triage, while the discussion candidly addresses data quality, ethical considerations, and real-world deployment challenges. Results show PanGu-based outputs outperform WenZhong across multiple metrics, though still fall short of human-level performance, underscoring the need for higher-quality data, improved architectures, and careful integration with professional care.

Abstract

The demand for psychological counselling has grown significantly in recent years, particularly with the global outbreak of COVID-19, which has heightened the need for timely and professional mental health support. Online psychological counselling has emerged as the predominant mode of providing services in response to this demand. In this study, we propose the Psy-LLM framework, an AI-based assistive tool leveraging Large Language Models (LLMs) for question-answering in psychological consultation settings to ease the demand for mental health professions. Our framework combines pre-trained LLMs with real-world professional Q\&A from psychologists and extensively crawled psychological articles. The Psy-LLM framework serves as a front-end tool for healthcare professionals, allowing them to provide immediate responses and mindfulness activities to alleviate patient stress. Additionally, it functions as a screening tool to identify urgent cases requiring further assistance. We evaluated the framework using intrinsic metrics, such as perplexity, and extrinsic evaluation metrics, with human participant assessments of response helpfulness, fluency, relevance, and logic. The results demonstrate the effectiveness of the Psy-LLM framework in generating coherent and relevant answers to psychological questions. This article discusses the potential and limitations of using large language models to enhance mental health support through AI technologies.
Paper Structure (40 sections, 2 equations, 10 figures, 7 tables)

This paper contains 40 sections, 2 equations, 10 figures, 7 tables.

Table of Contents

  1. Introduction
  2. Related Works
  3. Mental Health and Social Well-being in Overly Populated Cities
  4. Research Questions
  5. Research Scope
  6. Psy-LLM Framework
  7. Target Audience and Model Usage
  8. Large-scale Pre-trained LLMs
  9. PanGu Model
  10. WenZhong Model
  11. Collecting Large Scale Dataset
  12. Data Crawling
  13. Data Cleaning
  14. Data Analysis
  15. Model Training
  16. ...and 25 more sections

Figures (10)

  • Figure 1: The model layers and architecture of the PanGu model Zeng2021PanGuLA
  • Figure 2: The Query Layer in the PanGu Model
  • Figure 3: The number of characters in each sample
  • Figure 4: Word cloud of the frequent word within our dataset
  • Figure 5: Prepareing the training corpus
  • ...and 5 more figures