LAUSR

For the problem of knowledge overload in the process of online learning and the traditional algorithm's poor recommendation accuracy and real-time performance in the massive educational resources, a deep learning-based recommendation model for online educational resources is proposed. First, attribute features of learners and learning resources are extracted, and then text features of learning resources are extracted, and attention fusion of features at multiple different scales is performed using a multiscale fusion strategy. Finally, the fused features are used as input to the multilayer perceptron to train the classification model. Through testing a variety of educational resources, it is verified that the model in this paper has better real-time performance while maintaining high detection accuracy and outperforms the mainstream comparison model in several indexes, which have a certain application value. It provides a new way of thinking for educational platforms to build real-time educational resource recommendations.