Natural language understanding is the first step to realize human-machine dialogue.In the human-machine dialogue system,mature word segmentation dictionary ensures the correct segmentation of natural language and avoids the problem of semantic ambiguity caused by word segmentation ambiguity.However.,there are still two problems in word segmentation dictionary in human-machine dialogue system at present:1.the structure of word segmentation dictionary and the generalization of dictionary content;2... 

【文章來源】：華中師范大學(xué)湖北省 211工程院校 教育部直屬院校

【文章頁數(shù)】：78 頁

【學(xué)位級別】：碩士

【文章目錄】：
Acknowledgements
Abstract
Chapter 1 Introduction
    1.1 Research background
    1.2 Natural language understanding
    1.3 Problem raised
    1.4 Thesis structure and innovation points
Chapter 2 Overview of relevant theories
    2.1 Human-machine dialogue
        2.1.1 Basic concepts and development status of human-machine dialogue
        2.1.2 Four modes of human-machine dialogue
    2.2 Dictionary-based word segmentation technology
        2.2.1 The concept and development status of the word segmentation dictionary
        2.2.2 Dictionary-based word segmentation method
    2.3 Word segmentation dictionary
        2.3.1 Construction method of word segmentation dictionary
        2.3.2 Structure of the word segmentation dictionary in memory
Chapter 3 Comparative analysis of word segmentation performance
    3.1 Structure of the word segmentation dictionary
    3.2 Typical word segmentation dictionary structure
        3.2.1 Word segmentation dictionary mechanism based on the whole worddichotomy
        3.2.2 Word segmentation dictionary mechanism based on TRIE index tree
        3.2.3 Word segmentation dictionary mechanism based on each word dichotomy
        3.2.4 Dictionary mechanism based on double word hash index
    3.3 Dictionary analysis and comparison
        3.3.1 Experimental environment and data
        3.3.2 Conclusion
Chapter 4 Construction and optimization of subject dictionary
    4.1 Construction of double word hash word segmentation dictionary
        4.1.1 Dictionary function design
        4.1.2 Algorithm design of the dictionary
    4.2 Construction of domain corpus
    4.3 Optimization of subject dictionary
        4.3.1 Maximum inverse matching principle
        4.3.2 Parallel segmentation ambiguity resolution rules
    4.4 Comparison of experimental results
        4.4.1 Experimental data and evaluation criteria
        4.4.2 Analysis of results
Chapter 5 Application of subject dictionary in human-machine dialogue
    5.1 Definition of human-machine dialogue system
    5.2 Structure of human-machine dialogue system
    5.3 Implementation of human-machine dialogue system
    5.4 Analysis and comparison of results
Chapter 6 Summary and prospect
Reference
Appendix


【參考文獻】：
期刊論文
[1]基于Hash結(jié)構(gòu)的逆向最大匹配分詞算法的改進[J]. 丁振國,張卓,黎靖.  計算機工程與設(shè)計. 2008(12)
[2]面向信息檢索的自適應(yīng)中文分詞系統(tǒng)[J]. 曹勇剛,曹羽中,金茂忠,劉超.  軟件學(xué)報. 2006(03)
[3]基于二字詞位圖表的漢語自動分詞詞典機制[J]. 蔣斌,楊超,趙歡.  湖南大學(xué)學(xué)報(自然科學(xué)版). 2006(01)
[4]基于有向圖的雙向匹配分詞算法及實現(xiàn)[J]. 陳耀東,王挺.  計算機應(yīng)用. 2005(06)
[5]一種中文分詞詞典新機制——雙字哈希機制[J]. 李慶虎,陳玉健,孫家廣.  中文信息學(xué)報. 2003(04)
[6]基于知識的文本摘要系統(tǒng)研究與實現(xiàn)[J]. 孫春葵,李蕾,楊曉蘭,鐘義信.  計算機研究與發(fā)展. 2000(07)
[7]中文文本過濾的信息分流機制[J]. 林鴻飛,李業(yè)麗,姚天順.  計算機研究與發(fā)展. 2000(04)
[8]面向特定領(lǐng)域的理解型中文自動文摘系統(tǒng)[J]. 李蕾,鐘義信,郭祥昊.  計算機研究與發(fā)展. 2000(04)
[9]基于概念的文本結(jié)構(gòu)分析方法[J]. 林鴻飛,戰(zhàn)學(xué)剛,姚天順.  計算機研究與發(fā)展. 2000(03)
[10]基于篇章多級依存結(jié)構(gòu)的自動文摘研究[J]. 劉挺,王開鑄.  計算機研究與發(fā)展. 1999(04)



本文編號：3148112