摘要
互联网产生的海量信息带来了"信息超载"的问题。文章基于协同过滤算法对用户喜好进行了研究。阐述了协同过滤的基本思想,对用户喜好数据的采集及预处理过程进行了研究;在数据分析过程中提出几种常用的计算相似度的方法并进行了比较;研究了协同过滤算法的两个分支的不同适用场景,并与基于内容的算法进行比较,对现有算法存在的不足提出了改进。
The massive information generated by the Internet brings the problem of ＂information overload＂. This paper studies user preferences by using collaborative filtering algorithm; Describes the basic idea of collaborative filtering, and studies the acquisition and pretreatment of user preference data; in the process of data analysis, several common used similarity algorlthms are proposed and compared; the two branches of the collaborative filtering algorithm in different applicable scenes are studied, and compared with the content-based algorithm to improve the existing algorithms.
出处
《计算机时代》
2017年第7期56-59,共4页
Computer Era
关键词
协同过滤
用户喜好
数据采集
预处理
相似度
collaborative filtering
user preferences
data acquisition
pretreatment
similarity
