An Explainable Educational Resource Recommendation Model Based on Matrix Factorization
Xiaolin Gui, WU Fu-ying, Xiaoyan Liu, Yugen Yi, Zhenzhen Luo, Bing Li · 2021 7th IEEE International Conference on Network Intelligence and Digital Content (IC-NIDC) · 2021
The recommendation algorithm based on hidden variables are widely used in educational resource recommendation systems. However, such algorithms and their recommendation results lack explainability, which affects the application effect of recommendation. Therefore, we propose an explainable educational resource recommendation (EERR) model to solve this problem. The model is constructed by three steps. To begin with, we extract explainable features from educational resource manually. Then, the recessive feature is correlated with explicit feature by using of matrix decomposition. Finally, the alternating least square algorithm is used to obtain the recommended results. Experiment results show that the proposed model has better performance under the RMSE evaluation criteria, and it can improve users' trust in the recommendation system.