基于FCA的邮件分类系统模型  

Information Retrieval Model for Search Engine Based on FCA and Its Realization

作　　者：许研[1] 沈夏炯[1] 

机构地区：[1]河南大学计算机与信息工程学院

出　　处：《计算机与现代化》2007年第4期101-104,共4页Computer and Modernization

基　　金：河南省自然科学基金资助项目(0311011700)

摘　　要：对于大量的电子邮件,传统的分类方式是使用基于关键字比较的分类系统的过滤器对邮件进行分类,这种方式存在缺乏灵活性、分类粗糙等缺点。本文提出了一种基于形式概念分析(FCA)的邮件分类模型,在传统规则分类方法的基础上引入FCA分类机制,即利用FCA来抽取隐含在邮件文档中的潜在的概念关系,分析文档之间的相关性。用户对邮件文档的浏览是在一个根据概念划分的概念格结构上进行的。试验验证了该模型的可行性,试验结果表明该模型较好地解决了概括性的分类问题。For a large number of E-mails, traditional sorting fashion is to sort them by the filter of the sorting system which is based on keyword comparison. Aiming at actual fashion＇ s problem of returning rough sorts, this paper puts forward an E-mail sorting model based on Formal Concept Analysis（ FCA）, which adopts potential concept relation between the mails and analyzes rel- ativity between them by FCA. The user browses mails organized on Galois（concept） lattices. The paper emphasizes the implementation idea of E-mail sorting model. The test validates its feasibility, and test result indicates this model can make up the lim- itation of actual E-mail sorting system well.

关 键 词：FCA 属性抽取 邮件分类模型 邮件分类系统 

分 类 号：TP311[自动化与计算机技术—计算机软件与理论]