文档介绍：第 28 卷第 8 期计算机应用研究畅28 畅8
2011 年 8 月 Vol 畅2011No
Application Research puters Aug
推荐系统中显式评分输入的用户聚类方法研究
崔春生
(河南财经政法大学计算机学院, 郑州 450002)
摘要: 从推荐系统的矩阵稀疏性问题出发,在保证推荐的覆盖率基础上,借助社区密度的思想并通过矩阵约
简来提高推荐的准确率。运用归一化处理的用户评分矩阵对一致性的评分矩阵进行了改造,并借助社区的概念
通过矩阵调整得到了一种基于用户相似评分偏好的聚类分析结果;通过构造社区网络结构图得到社区密度,并
选择其中的最大值作为聚类中心,运用紧致与分离性效果函数验证了聚类结果的可靠性。该方法的提出,在不
损害已有信息的基础上降低了计算复杂度,通过的数据检验了该方法在一定程度上的正确性和有
Movielens
效性。
关键词: 推荐系统; 聚类分析; 社区网络; 稀疏矩阵
中图分类号: 311 文献标志码: 文章编号: 1001唱3695(2011)08唱2856唱03
TP A
: /. .
doi j issn
Aggregation method of mender systems with way of sparse
matrix reduction by sociality structure
唱
CUI Chun sheng
(School puter Science, Henan University of Economics Law, Zhengzhou 450002, China)
&
Abstract: , . 唱
Started in the, sparse matrix of mender systems and. ensured the mended coverage rate Based on the i,
dea of sociality density improved the accuracy. by matrix, reduction With the user, ratings matrix which had been normalized
transformed the consistency score matrix Meanwhile with the. concept of sociality got an aggregation analysis result based on唱
user similarity. score preferences through adjusting the matrix, Computed the sociality densities by constructing a
work diagram Selected the maximum. as the cluster centers and, used pact and separation effect functions to. validate唱
the reliability, of ag