计算机技术与发展2018，Vol.28Issue(5)：23-26,31,5.DOI:10.3969/j.issn.1673-629X.2018.05.006
融合信息熵和加权相似度的协同过滤算法研究
Research on Collaborative Filtering Algorithm Based on Information Entropy and Weighted-similarity
摘要
Abstract
Aiming at the problem of sparsity and only considering the number of common items to score between users and ignoring the specific score in traditional collaborative filtering,we propose a new algorithm model based on information entropy and weighted similari-ty,where we introduce the difference information entropy based on information entropy,which mainly consider the influence of specific score to common items,effectively alleviating the poor recommendation equality caused by sparse original rating matrix.In addition,by introducing the adjusted factor for weighted average for traditional similarity measurement method and the difference information entropy, we apply the weighted similarity to compute the similarity between the users and get better neighbors,to improve the precision of the neighbors.Last,it is applied to MovieLens dataset and compared with traditional collaborative filtering algorithm,and the experiments show that we can get better recommendation results,proving its effectiveness and feasibility.关键词
协同过滤/稀疏性/差异信息熵/加权相似度Key words
collaborative filtering/sparsity/difference information entropy/weighted-similarity分类
信息技术与安全科学引用本文复制引用
李玲,王移芝..融合信息熵和加权相似度的协同过滤算法研究[J].计算机技术与发展,2018,28(5):23-26,31,5.基金项目
国家自然科学基金(K13A300050) （K13A300050）
