计算机应用研究2017，Vol.34Issue(8)：2264-2268,5.DOI:10.3969/j.issn.1001-3695.2017.08.005
基于加权Slope One的协同过滤个性化推荐算法
Personalized collaborative filtering recommendationalgorithm based on weighted Slope One
摘要
Abstract
The traditional CF algorithm has the problem of cold start,data sparseness and low operation efficiency.This paper chose Slope One algorithm which was more efficient and accurate than traditional CF algorithm for research and analyses its advantages, principle and process.It pointed out shortcomings of Slope one algorithm that it did not take user interest changes and user similarity into account.This paper put forward improved scheme of Slope one algorithm based on user interest forgetting function and user nearest neighbors.The experimental test on MovieLens dataset proves the feasibility and better time performance of improved scheme.关键词
推荐算法/协同过滤/邻居选择/用户兴趣遗忘函数Key words
recommendation algorithm/collaborative filtering(CF)/neighbor-selection/user interest forgetting function分类
信息技术与安全科学引用本文复制引用
李桃迎,李墨,李鹏辉..基于加权Slope One的协同过滤个性化推荐算法[J].计算机应用研究,2017,34(8):2264-2268,5.基金项目
国家科技支撑计划资助项目(2014BAH24F04) （2014BAH24F04）
国家社科基金资助项目(15CGL031) （15CGL031）
中央高校基本科研业务费资助项目(3132016306) （3132016306）
