부산대학교 도서관

This paper proposes a novel personalized recommendation algorithm for learning resources based on differential evolution(DE) and graph neural networks(GNN). By representing learners and learning resources as graph data and incorporating a multi-head attention mechanism, we have developed an effective method for personalized recommendation. The differential evolution algorithm is utilized to optimize model hyperparameters, resulting in improved recommendation performance. We conducted experiments on a widely used personalized learning resource dataset, comparing our method with several classical recommendation algorithms. The results demonstrate significant advantages of our approach in terms of accuracy, recall, $\Gamma 1$ score, and RMSE value.