Innovative Exploration of Ideological and Political Education in Colleges and Universities in the Internet Era

Abstract

Abstract This article deeply explores the behavior and effect of online learning in ideological and political education in colleges and universities, firstly, it clarifies the mechanism of the occurrence of online ideological and political learning behavior, and constructs the corresponding indicators of learning behavior. Using CART tree and XGBoost model, the article ranks the feature importance of learning behaviors. It combines with Bayesian network to construct a comprehensive analysis model to explore the causal relationship between learning behaviors and learning effects. By analyzing the data of M online learning platform in 2021, the study found that resource learning features have the most significant impact on learning performance, especially the indicators of video viewing time, number of homework submissions and number of online discussions. The study results show that when learning resources are rich and professional, learning performance is significantly improved, providing an effective way to optimize the teaching quality of online Civics education.