基于概念图与WCN的二次语义检索的研究  

Research of Twice Semantic Retrieval Based on Conceptual Graph & WCN

作　　者：付跃华[1,2] 谢红薇[1] 余雪丽[1] 

机构地区：[1]太原理工大学计算机与软件学院,山西太原030024 [2]中国工商银行山西省分行,山西太原030024

出　　处：《微电子学与计算机》2009年第5期145-148,共4页Microelectronics & Computer

基　　金：山西省自然科学基金项目(20051035)

摘　　要：针对以用户目标为驱动的语义服务,提出了一种个性化的检索框架.采用加权概念网络表示用户模型,以概念图的形式定义需求中概念间的语义关联关系,充分表达用户的真实需求.给出了概念图与加权概念网络的匹配算法,对利用该算法得到的初步检索结果,考虑用户兴趣、职业特征及使用偏好等因素,进行二次排序,使得查询结果更富有个性化.实验结果表明,信息反馈的准确率得到了有效提高.For the semantic service driven by user＇s purpose,a personalized retrieval framework to its user is proposed.The weighted concept network（WCN）is adopted to represent the user profile.To express the actual requirements of users,conceptual graph is introduced to define the semantic relationship between concepts in query formally.The similarity algorithm between conceptual graph and WCN is proposed,too.The order of results is set twice before it is responded to users.The result set is more personalized.The experiment shows that the precision of the feedback has been improved efficiently.

关 键 词：语义检索 概念图 加权概念网络 个性化 

分 类 号：TP302[自动化与计算机技术—计算机系统结构]