基于向心扩散加权XML模型的异构用户个性化模式匹配方法  

The Matching Algorithm of Heterogeneous User Personalized Profile Based on Centripetal Spreading Weighted XML Model

作　　者：李树青[1] 刘晓倩[1] 

机构地区：[1]南京财经大学信息工程学院,南京210046

出　　处：《现代图书情报技术》2012年第5期32-40,共9页New Technology of Library and Information Service

基　　金：国家自然科学基金项目"基于通用加权XML模型的个性化用户兴趣本体研究"(项目编号:71103081);江苏省高校自然科学研究面上资助项目"通用加权XML模型在便携式个性化用户兴趣本体中的表达方法研究"(项目编号:11KJB630001)的研究成果之一

摘　　要：介绍一种利用同文词语共现和引文词语共现分析实现的领域本体自动构建方法,该本体采用加权XML模型,利用概念联系中的权值设定可以有效地表达用户兴趣程度的差异,并利用基于向心扩散的扩散激活方法对用户兴趣特征及其联系提供更强的表达能力,以便于发现更有价值的潜在用户兴趣。进而介绍如何利用该本体按照"先打碎后重构"的策略将异构用户个性化模式转换为可以进行比较的一致模式,并对相关的异构用户个性化模式匹配方法做出详细说明。最后总结相关测试实验及其结果。This paper introduces an automatic construction method for domain Ontology implemented by words co - oc- currence analysis in both document and citation. This Ontology adopts weighted XML model and uses weight in concepts and their relationship to express the difference of users＇ interest effectively, which can improve the ability of expressing users＇ interest and their relationship with centripetal weight spreading activation strategy in order to explore more valuable users＇ interest. Meantime, this paper also discusses how to use this Ontology to transform heterogeneous user personalized profile to consistent comparable model with the broken - and - reconstruction strategy, and how to match corresponding heterogeneous user personalized profile in detail. Finally, the result of correlative tests and experiments are concluded.

关 键 词：领域本体 加权XML 个性化 词语共现分析 

分 类 号：G202[文化科学—传播学]