基于概念格的查询扩展词推荐  

A Query Expansion Word Recommending Model Based on Concept Lattice

作　　者：崔骏[1] 张选平[1] 高由兵[1] 

机构地区：[1]西安交通大学电信学院软件研究所,陕西西安710049

出　　处：《微电子学与计算机》2008年第5期158-160,164,共4页Microelectronics & Computer

基　　金：国家自然科学基金项目(60673087);西安交通大学"九八五"项目

摘　　要：概念格是一种擅长描述层次关系的数学工具,在规则提取和数据分析中有广泛的应用.引入概念格理论对页面——概念形式背景建立了数学模型,在概念格基础上提出了一种查询扩展词生成算法.利用概念格Hasse图以及关联规则置信度以较高的效率生成扩展词作为二次搜索关键词,使信息搜索达到更好的效果.该算法在Diggol智能元搜索引擎上予以实现,取得了良好的效果.Concept Lattice is a kind of mathematic tools which is adept in describing hierarchy relationship, and is widely used in the rule extraction and the data analysis. This paper import concept lattice theory to build a mathematic model for the document-concept formal context. Using the concept lattice, Hasse diagram and confidence rate this article presents a query expansion word generation algorithm based on concept lattice, and provides more accurate similar terms as keywords to improve the retrieval effectiveness. This algorithm has been implemented on the Diggol intelligent meta-seareh engine, and has a good performance.

关 键 词：查询扩展 概念抽取 概念格 规则抽取 

分 类 号：TP391.11[自动化与计算机技术—计算机应用技术]