A Study of Teaching Interaction Strategies and Learning Effect Enhancement Based on Reinforcement Learning in Open Education Platforms

Abstract

In this paper, we first constructed an adult learner portrait and preprocessed the data for user behavior extraction. K-prototype and Pearson correlation coefficient based on the hybrid metric of Hamming distance and Euclidean distance are used to analyze the student learning behaviors and visualize them. Then in order to improve the learning effect of open education in adult education, a dialog strategy model based on user profiling and deep reinforcement learning is proposed, and comparative experiments are carried out on two datasets, ReDial and INSPIRED, to verify the effectiveness of the proposed algorithm. Finally, three classes in the open education platform in adult education are selected as research objects, and the interactive platform of teaching dialogue is applied to carry out experimental teaching, and the objective effect is analyzed by using the Spss mean difference test-t-t-test. The results show that compared with the mainstream algorithms, the method proposed in this paper is 0.453 in Recalla@50, 14.2% in Dist-3, 6.2% and 3.9% in BLEU-1 and BLEU-2. Secondly, the teaching platform established in this paper has a positive effect on the learning effect of open education.