Research on Artificial Intelligence Interaction Characteristics Affecting E-Learning Users' Sense of Trust

This paper mainly studies the reinforcement learning method with human-computer interaction function. In this project, a multi-agent cooperative optimization method based on multi-agents is proposed to realize the efficient retrieval of multi-agents in the system. Aiming at the problems of data sparsity and cold start in collaborative filtering algorithms, this project intends to study “online collaborative filtering algorithm based on confidence”. The evaluation data of students trusted by the researchers will be integrated so that it can more accurately reflect the preferences of students and identify students who share the same interests. The experimental results show that the reinforcement learning algorithm with human-computer interaction can greatly improve the learning efficiency and speed up the convergence of the learning process. The results of this study provide a possible method to apply human-computer interaction reinforcement learning technology to the study of complex tasks. This paper discusses how to integrate the two methods to make it better to complete complex tasks.