基于MMTD的中文副词模糊语义的量化研究  

Quantification Research on the Fuzzy Semantics of Chinese Adverbs Based on MMTD

作　　者：谢丁[1] 洪龙[1,2] 

机构地区：[1]南京邮电大学计算机学院,江苏南京210046 [2]软件开发环境国家重点实验室,北京100191

出　　处：《南京邮电大学学报（自然科学版）》2012年第4期48-52,共5页Journal of Nanjing University of Posts and Telecommunications：Natural Science Edition

基　　金：国家高技术研究发展计划(863计划)(2009AA043303);国家重点实验室开放课题(SKLSDE-2011KF-04)资助项目

摘　　要：自然语言理解是人工智能领域的核心课题,副词的模糊语义深刻影响着自然语言理解研究。针对副词在汉语语义中的作用,采用MMTD对其模糊语义进行了研究。根据英文副词频度数值研究的已有成果,建立了中文频度副词真值程度函数,在设定中文副词的程度数值后,建立了程度副词真值程度函数,提出副词连接概念及其量化方法。基于MMTD的中文副词的量化是语句语义数值化处理的前期准备,这些工作试图为自然语言理解建立一种新方法。Natural language understanding is a core subject in the artificial intelligence.The semantic fuzziness of adverbs plays a decisive role in the research of natural language understanding.This paper presents the research on the fuzzy semantics of Chinese adverbs using MMTD.According to the result of numerical value of English adverbs of frequency,this paper set up the truth degree function of Chinese adverbs of frequency.The truth degree function of Chinese adverbs of degree was established after setting their values.The concept of adverbs connection and the method of its quantification were both proposed.Quantification of the fuzzy semantics of Chinese adverbs based on MMTD is the preparation of the quantification research on Chinese sentences.These efforts attempt to build a new natural language understanding method.

关 键 词：副词 模糊语义 真值程度 量化 中文 

分 类 号：TP183[自动化与计算机技术—控制理论与控制工程]