论文部分内容阅读
利用用户兴趣可以有效地提高语义对等网环境下信息检索的效率,如何准确构建用户兴趣模型是关键。鉴于本地节点的信息资源可以有效反映用户兴趣,文章提出利用组织与管理本地节点资源的知识地图构建节点用户兴趣模型。主要思路是利用本体描述语言OWL描述本地知识实体及其关系,形成反映节点用户全局知识结构的知识地图,依据支持向量机分类原理从知识地图抽取出的兴趣特征训练集挖掘用户兴趣,最终形成用户兴趣模型并以兴趣描述文档的形式保存。
The use of user interest can effectively improve the efficiency of information retrieval in semantic P2P networks, and how to build a user interest model is the key. In view of local node information resources can effectively reflect the user interest, this paper proposes to construct node user interest model by using the knowledge map of organizing and managing local node resources. The main idea is to describe the local knowledge entities and their relations using ontology description language OWL and form a knowledge map that reflects the global knowledge structure of node users. According to the training set of interest features extracted from knowledge map based on support vector machine classification principle, the user interest is formed and users are finally formed The interest model is saved as an interest description document.