基于概念格标尺的广义匹配研究  

Augmented Pattern Matching Based on Concept Lattice Scale

作　　者：智东杰[1] 智慧来[2] 刘宗田[2] 

机构地区：[1]河南理工大学计算机科学与技术学院,焦作454150 [2]上海大学计算机工程与科学学院,上海200072

出　　处：《计算机科学》2008年第11期135-136,146,共3页Computer Science

基　　金：国家自然科学基金(项目批准号60575035)

摘　　要：为了提高计算机对自然语言的理解能力,实现概念的语义匹配,提出了基于概念格标尺的广义匹配。对现有的模式匹配的概念进行拓宽,利用概念格特有的层次关系建立语义标尺,通过标尺的同构来实现语义的匹配,同时分析了标尺刻度不一致时的刻度转换。研究表明该方法符合人们对现实世界中事物的认识,能够提高计算机对自然语言处理的智能。In order to improve computers＇ understanding ability of natural language, semantic pattern matching based on concept lattice scale was promoted. This method is seen as an augmentation of traditional pattern matching. It first utilized concept lattice＇s own hierarchical structure establishing semantic scale, then used isomorphs between two scales to realize semantic pattern matching of two different concepts. Graduation conversion between different scales was also studied. The research shows that our method confirms the way people look at the world, and can be realized by the computer to improve natural language processing ability.

关 键 词：概念格 标尺 模式匹配 映射 语义距离 

分 类 号：TP311.13[自动化与计算机技术—计算机软件与理论] TM934.17[自动化与计算机技术—计算机科学与技术]