基于上下文无关语法的英语句法剖析系统.docx

基于上下文无关语法的英语句法剖析系统摘要本毕业设计做的是基于上下文无关语法的英语句法剖析系统。本系统可分为语料库和程序处理两个模块。语料库模块由包含5492个考研词汇的词汇表、796个常用短语的短语表、非规则动词表、第三人称单数表、非规则名词表、规则表、标记集表以及对词汇表、常用短语表进行标记所产生的包含7359个单词标记的单词标记表和包含796个短语标记的短语标记表。其中非规则动词表、非规则名词表、第三人称单数表、规则表、标记集表以及单词标记表和短语标记表都是手工进行标记，输入所建立的。程序处理模块可分为单词识别、句子分解和句法剖析3个部分。单词识别部分是对输入的英文句子进行分词处理，然后对分离出的每个词在语料库中进行检索，对于变形后的单词要将其还原成原型然后再检索，检索成功则将所检索的单词进行词性标记，否则输出相应的处理信息。句子分解部分是对单词识别的结果予以分析，将其中的短语作为一个整体元素给予词性标记，最终将句中各元素及其标记以树状的形式显示出来。句法剖析部分是在前两个部分的基础上对用户所输入的句子进行句法结构的剖析，分析出句中各元素在句中所占的角色。若句法剖析成功则以树状的形式将用户所输入的句子的结构显示出来，否则输出相应的处理信息。关键字：自然语言理解语料库单词识别句子分解句法剖析Context-free grammar based Analysis system of English AyntaxAbstractThis projection is to do context-free grammar based analysis system of English syntax. The system can be divided into two modules.The first is corpus and the second is Program processing.There are some tables in the corpus,such as words table,phrase table,irregular verb form,the third person singular form,irregular noun form,rules table, mark set table,word type table,phrase type table.There 5492 words in the words table and 796 phrases in the phrase table.there are 7359 records in the word type table and 796 records in the phrase table.Word type table marked by hand,and so is phrase type table.Program processing module can be divided into word recognition,sentence decomposition and syntactic snalysis.The function of word recognition is to isolate words from the sentence which user input.The function of sentence decomposition is to isolate the independent element in the sentencewhich user input.Syntactic snalysis is on the basis of the first two steps,its function is to analysis of sentence which user input.If the result of analysis is right,output the result to the screen.Otherwise output the sentence can not be analyzed.Key words:Natural Language Understanding Corpus Word RecognitionSentence Decomposition Syntactic Analysis目录第一章前言31.1课题研究的背景知识31.2研究自然语言理解的意义3第二章系统概要设计32.1系统实现所基于的理论知识32.2系统实现功能概述32.3系统实现方法概述32.3．1语料库设计32.3．2程序设计构架3第三章系统详细