Knowledge Graph Construction and Personalised Learning Path Generation Mechanism for Online English Teaching and Learning

With the accelerated development of educational intelligence, the application of online English teaching has become more and more widespread. This paper constructs a knowledge map for online English teaching, which serves as the foundation for a personalized learning path generation model, with the goal of offering students deeply personalized English teaching services. Spectral convolution is used to transfer the filter and graph signals in the graph convolution network to the Fourier domain for computation, and the graph convolution neural network is used to generate the online English teaching knowledge map. Learner preferences are obtained by improving the knowledge graph convolutional network, and the learning path recommendation algorithm MKGLR is proposed. 50 first-year students of educational technology at Henan Normal University in China have been selected as the research subjects to carry out an online English teaching experiment. After the experiment, the average study duration of students decreased from 83.62 hours to 76.8 hours, the performance of online learning behavior improved significantly, and the number of classroom interactions, the number of homework submissions, and the number of resource accesses increased compared to the pre-experiment (Sig<0.01). There was a significant difference. The evaluation values of the usefulness and satisfaction dimensions increased by 43.42% and 26.77%, respectively, compared to the pre-experiment.