摘要
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 efficiency of information retrieval in the semantic peer-to-peer network can be effectively improved by the use of user interest,and the crux is how to accurately construct the user-interest model.In view of the fact that the information resources of the local peer can effectively reflect the user interest,this paper proposes to use the knowledge map,that organizes and manages the resources of the local peer,to construct a user-interest model.The main idea is as follows.Firstly,use the ontology description language OWL to describe the local knowledge entity and its relationship,forming the knowledge map that represents the user overall knowledge structure.Secondly,mine the user interest with the training set of the interest features,which is extracted from the knowledge map according to the classification principle of the support vector machine.Finally,form the user-interest model and save it as the user profile.
出处
《情报理论与实践》
CSSCI
北大核心
2012年第2期104-108,共5页
Information Studies:Theory & Application
基金
国家自然科学基金项目"基于语义对等网的知识组织和灵捷检索研究"(项目编号:70873090)
中央高校基本科研业务费专项资金"基于社区的P2P信息资源的语义检索模型研究"(项目编号:72104234)的成果
关键词
用户兴趣模型
语义对等网
知识地图
user-interest model
semantic P2P network
knowledge map
