201. Optimization path of campus mental health service system driven by educational big data

Abstract Background Campus mental health services face real challenges such as rapidly growing demand, lagging risk identification, and uneven allocation of intervention resources. Early identification and continuous monitoring of mental health problems are crucial for reducing the risk of symptom exacerbation and improving intervention effectiveness. With the development of educational big data technology, the comprehensive analysis of student behavior, learning, and psychological data offers new possibilities for optimizing the campus mental health service system. This study aims to explore the operational effectiveness of a campus mental health service system driven by educational big data and analyze its practical value in improving students’ mental state and enhancing their risk identification capabilities. Methods The study employed a quasi-experimental design, selecting 420 undergraduate students from two universities as research subjects. One university implemented a mental health service system supported by educational big data as the intervention group, while the other university used a conventional mental health service model as the control group. The educational big data-driven service system mainly includes a psychological risk early warning system integrating multi-source data, dynamic tracking of individual mental state, and a tiered intervention delivery mechanism. Mental state assessment used the anxiety and depression dimensions of the Symptom Checklist-90 (SCL-90), and psychological stress level was assessed using the Perceived Stress Scale (PSS). The intervention period was one semester, with assessments conducted at the beginning and end of the semester. Statistical analysis used paired t-tests and independent samples t-tests, controlling for variables such as grade and gender. Results End-of-term assessment results showed that the SCL-90 anxiety dimension score in the intervention group decreased from 2.12 ± 0.51 to 1.58 ± 0.47, while that in the control group decreased from 2.09 ± 0.49 to 1.92 ± 0.50, with a significant difference between the groups (t = 4.27, p<.001). The depression dimension score decreased from 2.06 ± 0.48 to 1.54 ± 0.45 in the intervention group, while the decrease in the control group was only from 2.03 ± 0.46 to 1.87 ± 0.47, with a statistically significant difference (t = 3.89, p<.001). The PSS score decreased significantly more in the intervention group than in the control group (t = 3.45, p=.001). Further analysis revealed a significant negative correlation between the number of risk warnings triggered by educational big data and the degree of improvement in psychological symptoms (r = -0.36, p<.001), suggesting that data-driven services have a positive role in psychological state regulation. Discussion A campus mental health service system driven by big data in education can effectively improve students’ anxiety and depression levels and enhance the timeliness of psychological risk identification and intervention. From a psychotherapy perspective, this model helps to achieve early detection and tiered intervention for mental health problems, compensating for the shortcomings of traditional service systems in dynamic monitoring and precise support. Future research can further incorporate longitudinal tracking data to assess the system’s impact on the prevention of serious mental health problems and long-term recovery of psychological function, while exploring sustainable optimization paths for campus mental health services within the framework of data ethics and privacy protection.