摘要
针对离散评分不能合理表达用户观点和传统协同过滤算法存在稀疏性等问题,借鉴年龄模糊模型,提出了梯形模糊评分模型。该模型将离散评分模糊化为梯形模糊数,考虑了评分模糊性和信息量,通过梯形模糊数来计算用户相似度,据此设计了协同过滤算法,并证明了该算法是传统协同过滤算法在模糊域的扩展。实验表明,该算法在数据稀疏且用户数远多于项目数时性能突出,并且算法运行时间远小于传统协同过滤算法。
In order to reflect the actual case of human decisions and solve the data sparseness problem of traditional collaborative filtering recommendation algorithm, a trapezoid fuzzy model based on age fuzzy model was proposed. In this model, crisp point was fuzzified into trapezoid fuzzy number and the fuzziness and information of users' grade was taken into account when calculating user's similarity by trapezoid fuzzy number. Based on this model, the user fuzzy similarity-based collaborative filtering recommendation algorithm was designed. The algorithm was proved to be an extension of traditional collaborative filtering algorithm in fuzzy fields. The experimental results show that, the proposed algorithm performs better when implemented in the sparse dataset with more user than item, and its running time is much less than traditional collaborative filtering algorithm.
出处
《通信学报》
EI
CSCD
北大核心
2016年第1期198-206,共9页
Journal on Communications
基金
国家重点基础研究发展计划("973"计划)基金资助项目(No.2012CB315901)
国家高技术研究发展计划("863"计划)基金资助项目(No.2011AA01AA103)~~
关键词
协同过滤
梯形模糊评分模型
模糊距离
模糊相似度
collaborative filtering
trapezoid fuzzy model
fuzzy distance
fuzzy similarity