To effectively evaluate the mental health status of college students, an intelligent method leveraging Internet of Things technology is proposed. This method incorporates maximum likelihood estimation into the Internet of Things framework, constructing an objective function for mining students’ mental health data. By analyzing the unique characteristics of students’ mental health data and calculating corresponding feature vectors, the system aims to extract vital data features. Furthermore, it considers the impact of educational programs on mental health as a two-dimensional vector, facilitating a comprehensive analysis. The method categorizes evaluation indicators based on quantitative values, establishing a refined evaluation index system for student mental health. Leveraging various indicator data, an evaluation model is created to assess the mental health level of college students and generate insightful results. Experimental findings validate the efficacy of this approach in capturing the evolving nature of college students’ mental health, providing valuable guidance for psychological counseling.