Intelligent Task Allocation Model for University Teaching Based on Knowledge Graph and Hierarchical Soft Actor-Critic Algorithm

The allocation of teaching tasks in colleges and universities is becoming more and more troublesome, mainly because students want more personalized teaching and the structure of courses is becoming more and more complex. In the past, we used some fixed rules to allocate teachers. This traditional method is difficult to ensure that everyone's workload is fair, and it is also difficult to arrange the most suitable teachers to the right courses, not to mention taking care of the teachers' own preferences, especially in the current situation of so many changes in the college environment. In order to solve these problems, our research proposes an intelligent allocation model that combines knowledge graphs and a reinforcement learning algorithm called Hierarchical Soft Actor-Critic (Hierarchical SAC) to dynamically adapt to the allocation of teaching tasks. The knowledge graph we use actually records the complex relationships between teachers and courses, including teachers' expertise, course requirements, and various teaching constraints, so that we can make decisions more clearly. And this hierarchical SAC algorithm can also dynamically optimize and match teachers and courses more reasonably. This matching process is carried out in layers, such as considering teachers' abilities, course requirements, and class time. To verify our model, we conducted experiments using real data from three universities, with a total of 30 teachers and 100 courses. The results showed that this new model not only significantly improved the fairness of teacher workload, but also significantly improved the matching between teachers and courses, and the satisfaction of teachers increased. Compared with traditional rule systems, genetic algorithms, and deep Q networks, our method is indeed more fair and efficient.