联结主义下‘能’的语义排歧研究
发布时间：2022-05-08 19:14
语义排歧（Word Sense Disambiguation）一直是自然语言处理（Natural Language Processing）和语义学研究领域一个非常重要的课题,几乎覆盖了各种自然语言处理系统,其中包括信息检索,机器翻译,关键词的提取,语音识别,文本分类和自动文摘。语言学家和计算机科学家在探索自然语言的歧义问题上做了大量的工作,取得了很大的成绩。但现有的对自然语言排歧的研究主要集中在语法、词典、简单的词汇词义层面上,对上下文、语义、语境、语用等知识和信息虽有涉及,但对这些信息的挖掘还相当有限。对于情态助动词这样语义更加模糊、对语境更为敏感的词类的语义排歧,目前尚未发现。语义排歧的研究对象是语言,将语言学的研究成果用到语义排歧中,将有利于打破语义消歧的瓶颈,推动其更深入的发展。特别是在情态意义方面,从《马氏文通》开始,语言学家对汉语情态助动词做了深入且全面的研究。这些研究成果为构建汉语情态动词的语义排歧模型的设想提供了充足的理论基础。另外,语义消歧的发展也将服务于语言学研究。情态助动词的自动语义排歧,能够实现情态助动词的语义的自动标注,从而为语言学家运用大规模语料库研究情态助...
【文章页数】：128 页
【学位级别】：硕士
【文章目录】：
摘要
Abstract
Chapter 1 Introduction
1.1 Background of this study
1.2 Objectives of the present study
1.3 Rationale of the study
1.4 Outline of the present study
Chapter 2 Literature Review
2.1 Approaches to word sense disambiguation
2.2 Word sense disambiguation studies abroad
2.3 Studies of WSD in China
2.4 Studies of Chinese modal verbs
Chapter 3 Methodology and Data Collection
3.1 Connectionism
3.1.1 Human and Artificial Neurons
3.1.2 Structure of Artificial Neural Network
3.1.3 Advantages of Artificial Neural Networks
3.1.4 The learning process
3.1.5 Back propagation network
3.2 Research method and data collection
Chapter 4 Sense Categorization of ‘能
4.1 Study of sense categorization on ‘能’
4.2 Sense categorization of ‘能’proposed by Li
4.2.1 Epistemic modality
4.2.2 Participant-internal modality
4.2.3 Participant-external modality
4.2.4 Deontic modality
Chapter 5 The Construction of the Neural Network Model for Word Sense Disambiguation of ‘能’
5.1 WSD Model
5.2 Construct two sense-tagged corpora of ‘能’
5.3 Feature extraction
5.4 Transfer linguistic features into vectors
5.4.1 Define the input vectors
5.4.2 Define the output vectors
5.5 Determine the number of nodes in hidden layer
5.6 Model constructing in Matlab
5.6.1 About Matlab
5.6.2 Construct neural network model in Matlab
5.7 Summary
Chapter 6 Further Study
6.1 Distribution of four senses of ‘能’in corpora
6.2 Co-occurrence of ‘能’with linguistic features
6.2.1 Co-occurrence of ‘能’with syntactic features
6.2.2 Co-occurrence of ‘能’with semantic features
6.3 The contribution of semantic features and syntactic features to word of ‘能'
6.4 Summary
Chapter 7 Conclusion
References
Appendix I
Appendix Ⅱ
Appendix Ⅲ
Appendix Ⅳ
Acknowledgements
作者简介
【参考文献】：
期刊论文
[1]基于可拓学理论的汉语词义消歧[J]. 卢志茂,刘挺,李生. 哈尔滨工业大学学报. 2006(12)
[2]基于多分类器决策的词义消歧方法[J]. 全昌勤,何婷婷,姬东鸿,余绍文. 计算机研究与发展. 2006(05)
[3]利用BP神经网络的中文词义消歧模型[J]. 何婷婷,谢芳. 华中师范大学学报(自然科学版). 2005(04)
[4]基于MDL聚类的无导词义消歧[J]. 陈浩,何婷婷,姬东鸿. 小型微型计算机系统. 2005(10)
[5]情态与汉语情态动词[J]. 朱冠明. 山东外语教学. 2005(02)
[6]情态动词“能”在交际过程中的义项呈现[J]. 王伟. 中国语文. 2000(03)
[7]基于神经网络的汉语口语多义选择[J]. 王海峰,高文,李生. 软件学报. 1999(12)
[8]助动词“想”和“要”的区别[J]. 张维耿. 语言教学与研究. 1982(01)
本文编号：3652221
【文章页数】：128 页
【学位级别】：硕士
【文章目录】：
摘要
Abstract
Chapter 1 Introduction
1.1 Background of this study
1.2 Objectives of the present study
1.3 Rationale of the study
1.4 Outline of the present study
Chapter 2 Literature Review
2.1 Approaches to word sense disambiguation
2.2 Word sense disambiguation studies abroad
2.3 Studies of WSD in China
2.4 Studies of Chinese modal verbs
Chapter 3 Methodology and Data Collection
3.1 Connectionism
3.1.1 Human and Artificial Neurons
3.1.2 Structure of Artificial Neural Network
3.1.3 Advantages of Artificial Neural Networks
3.1.4 The learning process
3.1.5 Back propagation network
3.2 Research method and data collection
Chapter 4 Sense Categorization of ‘能
4.1 Study of sense categorization on ‘能’
4.2 Sense categorization of ‘能’proposed by Li
4.2.1 Epistemic modality
4.2.2 Participant-internal modality
4.2.3 Participant-external modality
4.2.4 Deontic modality
Chapter 5 The Construction of the Neural Network Model for Word Sense Disambiguation of ‘能’
5.1 WSD Model
5.2 Construct two sense-tagged corpora of ‘能’
5.3 Feature extraction
5.4 Transfer linguistic features into vectors
5.4.1 Define the input vectors
5.4.2 Define the output vectors
5.5 Determine the number of nodes in hidden layer
5.6 Model constructing in Matlab
5.6.1 About Matlab
5.6.2 Construct neural network model in Matlab
5.7 Summary
Chapter 6 Further Study
6.1 Distribution of four senses of ‘能’in corpora
6.2 Co-occurrence of ‘能’with linguistic features
6.2.1 Co-occurrence of ‘能’with syntactic features
6.2.2 Co-occurrence of ‘能’with semantic features
6.3 The contribution of semantic features and syntactic features to word of ‘能'
6.4 Summary
Chapter 7 Conclusion
References
Appendix I
Appendix Ⅱ
Appendix Ⅲ
Appendix Ⅳ
Acknowledgements
作者简介
【参考文献】：
期刊论文
[1]基于可拓学理论的汉语词义消歧[J]. 卢志茂,刘挺,李生. 哈尔滨工业大学学报. 2006(12)
[2]基于多分类器决策的词义消歧方法[J]. 全昌勤,何婷婷,姬东鸿,余绍文. 计算机研究与发展. 2006(05)
[3]利用BP神经网络的中文词义消歧模型[J]. 何婷婷,谢芳. 华中师范大学学报(自然科学版). 2005(04)
[4]基于MDL聚类的无导词义消歧[J]. 陈浩,何婷婷,姬东鸿. 小型微型计算机系统. 2005(10)
[5]情态与汉语情态动词[J]. 朱冠明. 山东外语教学. 2005(02)
[6]情态动词“能”在交际过程中的义项呈现[J]. 王伟. 中国语文. 2000(03)
[7]基于神经网络的汉语口语多义选择[J]. 王海峰,高文,李生. 软件学报. 1999(12)
[8]助动词“想”和“要”的区别[J]. 张维耿. 语言教学与研究. 1982(01)
本文编号：3652221
本文链接：https://www.wllwen.com/wenyilunwen/yuyanxuelw/3652221.html
