摘要
提出一种基于用户兴趣度矩阵和改进粒子群优化的数字图书馆文献推送方法。从用户兴趣行为特征入手,首先构建用户对文献的兴趣度矩阵,然后通过文献聚类和最近邻查询,产生最优推荐列表。其中利用改进粒子群优化算法对文献聚类进行优化,进而能够提供更高效的文献推送服务。
This paper proposes a digital library literature recommendation method based on user interest matrix and improved particle swarm optimization.Starting with the behavioral characteristics of user interest,the paper first constructs a user interest matrix to literature,then establishes an optimum recommendation list by literature clustering and the nearest neighbor searching.The paper uses the improved particle swarm optimization method to optimize literature clustering,which can provide a more effective literature recommendation service.
出处
《情报理论与实践》
CSSCI
北大核心
2012年第7期92-95,共4页
Information Studies:Theory & Application
关键词
文献推送方法
数字图书馆
应用研究
literature recommendation method
digital library
application study
