Effectiveness of Mental Health Services for Improving Minors’ Behavior in Community Corrections Using Online Platforms

The use of a network platform for the online correction of minors helps to optimize the correction effect. In this paper, we build a network platform for minor community correction, employ TAN Bayes to measure the corrective mental state, and utilize support vector machines to extract corrective content and identify relationships between it after text preprocessing. Next, the feature values are obtained by reducing the number of dimensions using probability matrix decomposition. The gradient descent method is then used to optimize the matrix. The feature matrix is then created, and the improved PMF is added to the ConvMF model. Finally, the Gray Wolf algorithm is used to improve the corrective content recommendation. After the platform construction was completed, the effect of applying this paper’s platform to two corrected minors was analyzed, and it was found that in the group of minors’ community correctional subjects, the correctional content recommended by the three different types of this paper’s platforms would all have different impacts in different dimensions compared to the whole. Before beginning corrective action using this article’s online platform, Case A’s evaluation on the 10 questions averaged 93.41 points, indicating a poor corrective psychological state and a high risk of mental health. After six months of online correction on this paper’s online platform, the psychological state improved significantly, with an average of 48.3 points on the posttest and a reduction of more than 40 points on each topic item. Compared to before the start of correction, the psychological risk has significantly decreased. Case B scored an average of 93.92 points on the 10-topic item pre-test, indicating an unacceptable mental health status that necessitates immediate intervention. After six months of online correction, On this paper’s online platform, the average posttest score was 47.69, with an average decrease of 46.23 points for each item, and significant results were achieved. This paper provides useful exploration for optimizing the method of community correction for minors and improving its effectiveness for minors.