Recommendation System for Ideological and Political Education Course Resources Based on Alternating Least Squares

In modern era, ideological and political course resources recommendation systems play an essential role in providing personalize course recommendations based on the student preferences. However, several traditional methods faced limitations of time sensitive rating relevance and computational inefficiency. To overcome these issues, Alternating Least Squares (ALS) algorithm is proposed, which recommends ideological and political courses by optimizing the student course interaction. Initially, the student course interaction matrix is generated using internal student engagement data. After that, ALS algorithm is utilized to factorize the sparse interaction matrix to handle the missing values. Next, student preferences are assessed by utilizing Analytic Hierarchy Process (AHP) and behavioral data is converted to preference values. Furthermore, the recommendation system is modelled as a Markov Decision Process (MDP), where student ratings evolve over time. Lastly, policy-based reinforcement learning is employed to the recommendation system and generated preference values are fed to the model for personalized and effective ideological and political course resources recommendations. The proposed ALS achieved better results in terms of Mean Absolute Error (MAE) and Root Mean Square Error (RMSE) with 0.50 and 0.97 values respectively and outperformed the existing Singular Value Decomposition (SVD) approach.