基于语义计算的中文歧义字段消歧算法  

Disambiguation Algorithm for Chinese Word Segmentation Based on Semantic Calculation

作　　者：邓凡[1] 鱼滨[2] 

机构地区：[1]西北大学信息科学与技术学院,陕西西安710069 [2]西安电子科技大学计算机学院,陕西西安710071

出　　处：《计算机技术与发展》2008年第6期107-110,共4页Computer Technology and Development

摘　　要：针对中文中歧义字段对中文处理及理解带来的诸多问题提出了一种基于自然语言理解的中文汉字歧义消除算法。对于交集型歧义和组合型歧义,利用《知网》为主要语义资源,以知识图知识表示方法,通过提出的字段消歧算法,对歧义字段以及上下文的语义进行计算,从而选出正确的句子切分方案,达到消除歧义的目的。经过实验数据表明本算法提高了中文歧义字段歧义切分的正确率。In this paper, based on a study of natural language understanding of Chinese characters algorithm to eliminate ambiguity. Aiming at cross ambiguity and combinational ambiguity, using the HowNet as the main semantic resources, using knowledge map＇s knowledge as representation, using the algorithrus mentioned in this paper based on semantic calculations, and calculating the discrepancy field of semantic context, to select the correct sentence cut programs, so as to achieve the objective to remove ambiguities. Experiments show that the algorithm does indeed contribute to improving ambiguity cut to the correct rate.

关 键 词：自然语言理解 交集型歧义 组合型歧义 词义消歧 关联度 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]