臺灣博碩士論文加值系統

對於非母語學習者而言，學習第二語言(Second Language)的同義詞(Synonym)用法並非是一件容易的事，這是因為即使是意義相同或相近的詞彙，其用法上卻未必相同。因此，第二語言學習者即使知道詞彙的意義，但在表達或寫作上往往可能發生用錯詞彙的情形，例如：{strong, powerful}兩個字雖然意義相近，但要表達”濃咖啡”的時候必須使用strong coffee，而非powerful coffee。有鑑於此，本研究之目的即在區別英語同義詞之用法，期望未來學習者能透過電腦輔助的方式瞭解同義詞的使用時機及語法結構。
本研究使用資訊檢索(Information Retrieval)領域常用的向量空間模型(Vector Space Model, VSM)來區別同義詞之用法。在實驗中，我們比較向量空間模型與兩種非監督式學習的方法：N連詞(N-gram)模型與點式交互資訊(Pointwise Mutual Information, PMI)，實驗結果顯示，向量空間模型獲得較佳的正確率。

For non-native learners, distinguishing synonyms of second language is not easy. The reason is that near-synonyms may have their specific usage and syntactic constraints. Therefore, second language learners may use wrong words when composing a text even though they know the meaning of the words. For instance, although strong and powerful have similar meaning, "strong coffee" is a collocation and "powerful coffee" is an anti-collocation.
This study proposes the use of a vector space model (VSM), which has been widely used in Information Retrieval (IR) community, to distinguish among near-synonyms. The VSM is compared to two unsupervised methods: n-gram and pointwise mutual information (PMI). Experimental results show that the VSM achieved higher accuracy that both n-gram and PMI.

書名頁.....i
論文口試委員審定書.....ii
授權書.....iii
中文摘要.....iv
英文摘要.....v
誌謝.....vi
目錄.....vii
表目錄.....viii
圖目錄.....ix
第一章 緒論 1
1.1 研究背景與動機 1
1.2 研究目的 2
1.3 研究方法 3
1.4 研究架構與流程 4
第二章 研究方法 5
2.1 點式交互資訊(Pointwise Mutual Information, PMI) 5
2.2 N連詞模型(N-gram) 6
2.3 向量空間模型(Vector Space Model, VSM) 8
第三章 實驗結果與分析 11
3.1 實驗設計 11
3.2 實驗結果 13
第四章 結論與未來展望 17
參考文獻 18

1.D. Pearce. 2001. Synonymy in Collocation Extraction. In Proceedings of the Workshop on WordNet and Other Lexical Resources at NAACL-01.
2.C. Fellbaum. 1998. WordNet: An Electronic Lexical Database. MIT Press, Cambridge, MA.
3.H. Rodriguez, S. Climent, P. Vossen, L. Bloksma, W. Peters, A. Alonge, F. Bertagna, and A. Roventint. 1998. The Top-Down Strategy for Building Eeu-roWordNet: Vocabulary Coverage, Base Concepts and Top Ontology, Computers and the Humanities, 32:117-159.
4.Dong, Z., & Dong, Q. 2006. HowNet and the Computation of Meaning. New Jersey: World Scientific Publishing.
5.L. C. Yu, C. H. Wu, A. Philpot, and E. H. Hovy. 2007. OntoNotes: Sense Pool Verification Using Google N-gram and Statistical Tests, In Proceedings of the OntoLex Workshop at the 6th International Semantic Web Conference (ISWC-07), Busan, Korea.
6.D. Inkpen. 2007. Near-Synonym Choice in an Intelligent Thesaurus. In Proceedings of Human Language Technologies: The Annual Conference of the North American Chapter of the Association for Computational Linguistics (NAACL/HLT-07), pages 356-363.
7.D. Inkpen. 2007. A Statistical Model for Near-Synonym Choice. ACM Trans. on Speech and Language Processing, 4(1):article 2.
8.M. Gardiner and M. Dras. 2007. Exploring Ap-proaches to Discriminating among Near-Synonyms, In Proceedings of the Australasian Technology Workshop, pages 31-39.
9.D. McCarthy. 2002. Lexical Substitution as a Task for WSD Evaluation. In Proceedings of the SIGLEX/SENSEVAL Workshop on Word Sense Disambiguation at ACL-02, pages 109-115.
10.I. Dagan, O. Glickman, A. Gliozzo, E. Marmorshtein, and C. Strapparava. 2006. Direct Word Sense Matching for Lexical Substitution. In Proceedings of the 21st International Conference on Computa-tional Linguistics and 44th Annual Meeting of the Association for Computational Linguistics (COL-ING/ACL-06), pages 449-456.
11.G. Salton and M. J. McGill, Introduction to Modern Information Retrieval. McGraw-Hill, New York, 1983.
12.Baeza-Yates, R. and B. Ribeiro-Neto. 1999. Modern Information Retrieval. Addison-Wesley, Reading, MA.
13.K. Church and P. Hanks. 1991. Word Association Norms, Mutual Information and Lexicography. Computational Linguistics, 16(1):22-29.