Collaborative Filtering Recommender Based on Feature Clustering and Time Series
This paper proposes a Collaborative filtering recommender based on feature clustering and time series,which uses K-means algorithm to cluster fault features to improve the calculation efficiency of similarity comparison.According to the multimodal fault feature fusion matching maintenance scheme,time series analysis is performed on the results,and the weight of the maintenance scheme is adjusted to ensure the accuracy and timeliness of the results.In addition,the mean strategy was used to achieve score fusion and establish a group recommender under multiple fault features.The experiment shows that compared with the UBCF and IBCF methods,the accuracy of our method has been improved by 1.27%and 4.46%and the recall has been improved by 2.77%and 0.61%.
recommendercollaborative filteringdeature clusteringtime series
NETL
NSTL
万方数据
