摘要
提出一种基于概念格的图书协同推荐方法。首先建立用户与图书之间的概念格,然后根据概念格中图书节点之间的继承关系,从下向上回溯用户之间的共同偏好,快速计算用户之间的相似度,以确定目标用户的"最近邻居"。最后通过计算协同推荐值,实现对目标用户推荐其可能感兴趣的图书,从而为图书的协同推荐提供支持。
The method of effectively recommending users books based on concept lattice is proposed in the process of collaborative recommendation of digital library. Firstly, the concept lattice of customers and books is built. Secondly, according to the inheritance relationship of nodes in concept lattice, tile similarity degrees between users can be calculated by searching the common preference between users, and then ＂nearest neighbors＂ of the target user can be confirmed. Finally, recommending books can be realized by calculating the collaborative recommendation value.
出处
《图书情报工作》
CSSCI
北大核心
2012年第17期131-135,共5页
Library and Information Service
关键词
数字图书馆协同推荐概念格相似度
最近邻居
digital library collaborative recommendation concept lattice similarity degree nearest neighbor
