摘要
协同过滤算法是当今推荐系统中最流行一种算法,可以细分为基于用户的协同过滤算法和基于物品的协同过滤算法,核心是寻找用户或物品之间的相似度。基于此,在传统算法求解相似度的基础上,提出了一种新的算法求解相似度,并通过实验验证了所提算法的优越性。
Nowadays,the collaborative filtering is one of the most popular algorithm in recommendation system while the collaborative filtering can be divided into user based and item based.But the key to this algorithm is to find the similarity.Based on the traditional way count the similarity,this paper proposes a new algorithm to count the similarity and proofs its advantage through some test.
作者
张梓培
Zhang Zipei(Guangdong University of Technology,Guangzhou Guangdong 510006,China)
出处
《信息与电脑》
2018年第21期51-53,共3页
Information & Computer
关键词
推荐系统
LCS
协同过滤
recommendation system
LCS
collaborative filtering