Design of Teaching Resource Recommendation Platform Based on Fuzzy Recommendation Algorithm in the Context of Education Informatization

With the popularity of online education platforms, the quantity and variety of teaching resources have rapidly increased. Due to the abundance and complexity of these teaching resources, an efficient recommendation system is needed to help users quickly find useful teaching content. In order to improve the updating speed and utilization efficiency of online teaching resources, the study proposes a teaching resource recommendation platform system combining Takagi-Sugeno fuzzy model and neuromatrix decomposition model to solve the lack of linear relationship in fuzzy recommendation algorithm. The results show that the Area Under Curve of the proposed model is 0.940, while the Area Under Curve of the graph neural network model is the smallest, only 0.892. The reason is that the graph neural network model cannot fully capture the correlation information between users and educational resources, leading to a decrease in recommendation performance. The accuracy of the model proposed within 0-500 iterations is 95.60%, which is 4.51% higher than the integrated model of convolutional neural networks and bidirectional long short-term memory networks. This indicates that the proposed model has good application effect and feasibility, achieving good teaching resource recommendation results. This has practical application value for improving the efficiency of recommending educational resources.