Optimization Strategies of Bayesian Modeling Algorithms for Multilingual Teaching Systems in Southeast Asian Universities

In the context of globalization, the development of multilingual teaching system has become a hot topic nowadays. Aiming at the multilingual teaching scenario, this paper proposes a Bayesian knowledge tracking method based on genetic algorithms for multi-knowledge points. The method adjusts the model structure of traditional Bayesian knowledge tracking, which can track students’ mastery of multiple knowledge points at the same time, and introduces a genetic algorithm for parameter optimization. Comparison with the standard Bayesian tracking model reveals that the experimental results of this paper’s method are superior in terms of accuracy, AUC value, RMSE value, and loss rate, possessing higher accuracy and lower loss rate. Conducting teaching experiments in multilingual teaching in a Southeast Asian university, the method achieved 75.8% accuracy in students’ knowledge tracking status, with an overall satisfaction score of 4.4 or more, and the difference in students’ performance between the two classes was significant as the application went deeper (p < 0.05). The Bayesian knowledge tracking model proposed in this study has achieved good results in assessing the accuracy of learners’ knowledge status, improving learners’ learning outcomes, and satisfaction with the application of the model.