DEVELOPMENT OF A YOUTH RECIDIVISM RISK ASSESSMENT SYSTEM USING MACHINE LEARNING UNDER THE FRAMEWORK OF THE FOUR NOBLE TRUTHS

Abstract

The development of a youth recidivism risk assessment system integrating the Four Noble Truths with Machine Learning technology to address the high recidivism rate among Thai youth, which reaches 41% within three years after release. The developed system has a structure aligned with the Four Noble Truths, beginning with problem identification (Dukkha) through empirical data collection, followed by causal analysis (Samudaya) using Machine Learning techniques such as Markov Chain algorithms and Unsupervised Learning to identify risk factors leading to recidivism. The system sets clear goals (Nirodha) to reduce recidivism rates and establishes solution approaches (Magga) that apply Explainable AI (XAI) and Fairness-Aware Scoring Systems for transparency and equity. System testing results show a high accuracy rate of 78% in identifying at-risk groups, leading to the design of effective individualized rehabilitation plans that reduced recidivism rates in the experimental group by 32% compared to the control group. The integration of Buddhist philosophy with modern technology not only enhances the efficiency of risk assessment and screening systems but also strengthens operational alignment with Thai cultural principles and thinking. This approach prioritizes the ultimate goals of rehabilitation and prevention over punishment, establishing a crucial foundation for sustainable development of the juvenile justice system.