通过句法位置提取中文关键词的实验研究

出版时间：2011-10  出版社：苏州大学出版社  作者：王家钺  页数：149  

内容概要

　　《独秀外国语言文学博士文库：通过句法位置提取中文关键词的实验研究》详尽描述了自然语言处理尤其是信息检索的各种主要方法，并对信息检索领域“相关性”这一核心概念进行了回顾与综述，在此基础上提出了基于句法位置提取关键词的初步设想，通过小型问卷调查证明了基本名物性短语的重要性。在实验部分，作者以从某技术型网站搜集的小型文本库为实验对象，首先证明了不同文本位置以及不同句法层级上产生关键词的可能性有显著差异。随后使用向量空间模型（VSM）提取其关键词，然后对文本库中的基本名物性短语的句法位置进行手工标注，通过这些位置提取关键词，再将二者进行对比。实验表明，通过句法位置提取的关键词与通过VSM提取的关键词没有显著差异。最后作者对这些实验结果进行了讨论。

作者简介

　　王家钺，广西师范大学外国语学院英语语言文学系主任、博士、副教授、硕士研究生导师。从事应用语言学、语料库语言学、计算机辅助翻译等领域的研究。在《外语界》《现代外语》等刊物上发表论文多篇，主持省厅级课题2项。

书籍目录

Chapter 1 Introduction1.2 Discussions on KE methodsChapter 2 Information retrieval: the ultimate goal2.1 Introduction2.3.1 Free text string searching2.4.1 The Vector Space Model2.4.2 Probabilistic approaches2.4.3 Document classification and clustering2.5 Linguistic approaches to IR2.5.1 Problems with statistical methods2.5.2 Non-statistical methods2.7 Nominal phrases2.7.1 Significance2.7.2 Phrase detection and extraction2.7.3 Base NPs2.8 SummaryChapter 3 Web search and relevance: trigger and rationale3.1 Web search: a good landing for IR discussions3.2 The "hypertext challenge"? 3.3 Search engines: performance and problems3.4 IR and relevance3.5 Relevance studies3.5.1 Anatomy of the concept3.5.2 Relevance assessment variation 3.5.3 System-oriented relevance ,3.5.4 User-oriented relevance3.5.5 System-oriented definitions of relevance3.6 Discussion and re-definitionChapter 4 Keyword extraction: methodology and practices4.1 KE: a good aid for IR4.2 What is "topic"? 4.3 KE review 4.3.1 Text segmentation and topic extraction……Chapter 5 ExperimentsChapter 6 Conclusions and further researchBibliographyAppendices

章节摘录

　　The English version is a nominal construction， whereas the Chinese version is obviously a verb construction. However， we suspect that this case is an exception， because the two versions do not semantically correspond exactly， and as a matter of fact "试析" （ "tentatively analyze" ） is used quite stereotypically in cases where the meaning of the title involves some " analytic"　work，　thus naturally having a greater likelihood of being used.　　The above examples may show how the same authors would represent their subject in two different languages， particularly in English and Chinese. To investigate people's response to titles of other authors， we also did an informal survey， in which a small number of graduate （Ph. D. and M. A.） students majoring in linguistics and applied linguistics， unaware of the purpose of our survey， were asked to give their Chinese translations for some selected article titles and book titles in English. （See the corresponding Appendix A for the questionnaire.） In this questionnaire， six of the original titles were verb phrases headed by a gerund; two were prepositional phrases headed by on or towards， and one title was a nominal phrase headed by "A holistic survey of..." （the same title as the example given a little earlier）. The following table shows the categorical shifts during their translation.　　……