Online Learning Resource Recommendation by Using Parallel Spectral Clustering Algorithm Based on Spark Platform
To improve the accuracy of online learning resource recommendation,spectral clustering was used to classify learning resources,recommend resources with high category similarity to users,a parallel spectral clustering algorithm based on Spark platform was proposed to enhance the efficiency of resource recommendation.First,the online learning resources and user features were extracted and initialized,then a spectral clustering model was established,and the vertex similarity matrix and normalized Laplacian matrix of the undirected graph on Spark platform were solved.Then the normalized segmentation for subset division was used,the class vector was solved by the normalized cut set optimization method,and the feature points of the class vector by line was output.Finally,k-means algorithm was used to cluster the feature points and obtain the clustering results.The experimental results show that compared with the common online learning resource recommendation algorithms,the spectral clustering algorithm based on Spark platform has higher recom-mendation accuracy and coverage,and has higher adaptability in real-time recommendation of massive learning resources.
NETL
NSTL
万方数据
