摘要： 以数字图书馆领域个性化服务为例,以空间向量模型表示用户兴趣和资源特征,并借助于构建的领域本体和“知网”知识词典对向量进行概念上的扩展,形成用户和资源特征概念空间向量,并通过向量相似度计算寻找最优的资源,从而为用户提供个性化信息服务。实验数据说明了基于概念的相似度计算比基于关键词的相似度计算具有更明显的优越性。

关键词: 个性化服务, 本体论, 特征提取, 空间向量, 相似度计算

Abstract: This paper proposes a new method to implement personalized service and introduces it in the example of digital library service. Users profile and resource features are represented by space vector model. And then these vectors are extended in conceptual level in virtue of domain ontology and Hownet dictionary. So conceptual space vector of users and resources are generated and can be used to find optimum resource for different user through computing their similarity. Therefore personalized service provides to users. Experimental data illustrates that the similarity based on concept is more efficient than the similarity based on keyword.

中图分类号: 

• TP391.3