Deep Reinforcement Learning-based Recommendation System for Ideological and Political Curriculum Learning Resources

Political and Ideological teaching is a social practice that uses political views, moral norms, and ideas to improve students' ideology and conform to social norms for Ideological and political curriculum learning resources. Traditional approaches to ideological and political education are ineffective because no proper personalized recommendations exist. Therefore, adapting to individual learners' needs, interests, and learning styles is difficult, and cold start challenges arise once a novel user joins a recommendation system without any historical data. In this paper, Deep Reinforcement Learning (DRL) is proposed to recommend mobile political and ideological education. Initially, the data is taken from the edX dataset followed by content extraction including recommendations based on historical records. Then, the feature selection with Recursive Feature Elimination (RFE) is used to select optimal recommendations from content. After that, the DRL is introduced to determine the selected features based on the chi-square value with the help of the chi-square test. Finally, the recommendation algorithm evaluates the value of the chi-square with corresponding recommendations generated from content extraction and recommends better mobile political, and ideological education. From the results, the DRL gives more accurate results in contrast with the existing Analytic Hierarchy Process (AHP) model in terms of accuracy (98.22%), recall (89.90%), and f1-score (95.45%) respectively.