Student Mental Health Support System Based on Multimodal Emotional Interaction

心理健康是社会普遍关注的问题，我国抑郁症患者群体人数持续扩大，发病群体呈现年轻化趋势，且高校学生群体占比持续升高，利用人工智能技术赋能高校心理健康工作刻不容缓。本文针对现有高校心理健康工作存在集中评测精度不足、隐患排查时效性差、传统面谈覆盖面窄等问题，提出了基于多模态情感交互的学生心理健康支持系统。该系统依托校园行为大数据，构建学生异常情绪及行为监测与预警机制；通过自主研发的大语言模型，实现学生情绪的动态识别与智能评估，并结合心理学理论，动态适配个性化疏导策略，支持多角色的情感陪伴与心理支持。系统在实际应用中表现出良好的效果，心理状态评测准确率超过85%，显著提升了心理服务的精准性与响应效率，为校园心理健康教育体系的智能化与科学化建设提供了有力的技术支撑。Mental health is a common concern in society. The number of depression patients in my country continues to expand, the incidence group shows a trend of younger age, and the proportion of college students continues to increase. It is urgent to use artificial intelligence technology to empower college mental health work. Aiming at the problems of insufficient centralized evaluation accuracy, poor timeliness of hidden danger investigation, and narrow coverage of traditional interviews in existing college mental health work, this paper proposes a student mental health support system based on multimodal emotional interaction. Relying on campus behavior big data, the system builds a monitoring and early warning mechanism for students’ abnormal emotions and behaviors; through the independently developed large language model, it realizes the dynamic recognition and intelligent evaluation of students’ emotions, and combines psychological theory to dynamically adapt personalized counseling strategies to support emotional companionship and psychological support for multiple roles. The system has shown good results in practical applications, with an accuracy rate of more than 85% in psychological state evaluation, which significantly improves the accuracy and response efficiency of psychological services, and provides strong technical support for the intelligent and scientific construction of the campus mental health education system.