我們提出了一個偵測錯誤與改正錯誤的方法。根據劍橋學習者語料庫 (Cambridge Learner Corpus) 的統計，在英文學習者常犯的錯誤中，除了拼字錯誤外，最多的就是動詞誤用 (Replace Verb Error)。我們使用了維基百科的編修紀錄來建立通道模型 (Channel Model) ，並結合搭配詞組 (Collocation) 來取出候選動詞組，用以來取代句子中可能錯誤的動詞。我們的改錯方法包含了句子剖析 (Parsing)、動詞辨認、應用迴歸分析於通道模型、 擷取候選詞組，最後重新排序候選詞組中的動詞，來達成動詞改正建議的目的。

According to the analysis of a sample of Cambridge Learner Corpus, the most frequent type of grammatical errors is using a wrong verb. We introduce a method for learning to detect and correct potential verb errors in a given sentence. In our approach, alternative verbs are considered to replace the verb based on verb- object collocations and a channel model trained on an error-annotated corpus. The method involves parsing the sentence, identifying the verbs, applying regression on channel model, retrieving a list of alternative verbs, and re-ranking each alternative. Our method combine channel and language models, resulting in high recall and precision rates in detecting and high MRR in correcting verb misuse.

Abstract ii
Acknowledgements iii
Contents iv
List of Figures vi
List of Tables viii

Chapter 1 Introduction ........................ 1
1.1 TheoreticalBackground ........................ 2
1.2 Motivation................................ 2
1.3 ThesisOrganization........................... 4

Chapter 2 Related Work ........................ 5
2.1  GEC for Different Error Types..................... 5

2.2  GEC Approaches ............................ 6

2.2.1 The Statistical Machine Translation. . . . . . . . . . . . . . 6
2.2.2 The Classification Approach.................. 7
2.2.3 HybridApproach ........................ 8

Chapter 3 Methodology ........................ 10
3.1  ProblemStatement........................... 10
3.2  Tagging Learner Corpus with Error Types . . . . . . . . . . . . . . 11
3.3  Applying Regression Model ...................... 13
3.4  Detecting and Correcting RV Errors.................. 17
3.4.1 IdentifyingVerbs ........................ 18
3.4.2 GeneratingCandidateList................... 20
3.4.3 Detecting RV Errors based on Language Model . . . . . . . 23
3.4.4 Re-rankingAlternatives .................... 24 

Chapter 4  Experimental Setting ........................ 26
4.1  Dataset ................................. 27
4.2  Correction Systems Compared..................... 31
4.3  Evaluation Metrics ........................... 32
4.3.1 Performance of RV Error Detection . . . . . . . . . . . . . . 32
4.3.2 Performance of Verb Suggestion................ 33 

Chapter 5  Results and Discussion ........................ 34

5.1 Results for RV Error Detection .................... 34
5.2 Results for Verb Suggestion ...................... 35
5.3 Error Analysis.............................. 38 


Chapter 6 Conclusion and Future Work .............................. 41

Reference .............................. 43

Appendices .............................. 47

Aoife Cahill, Nitin Madnani, Joel Tetreault, and Diane Napolitano. Robust sys- tems for preposition error correction using Wikipedia revisions. In Proceedings of NAACL-HLT, pages 507–517, 2013.

Daniel Dahlmeier and Hwee Tou Ng. Grammatical error correction with alter- nating structure optimization. In Proceedings of the 49th ACL, pages 915–923, 2011.

Rachele De Felice and Stephen G Pulman. A classifier-based approach to prepo- sition and determiner error correction in L2 English. In Proceedings of the 22nd Cocling, 2008.

Mariano Felice, Zheng Yuan, Øistein E Andersen, Helen Yannakoudakis, and Eka- terina Kochmar. Grammatical error correction using hybrid systems and type filtering. In Proceedings of the 18th CoNLL: Shared Task, pages 15–24, 2014.
Christiane Fellbaum. WordNet. Wiley Online Library, 1998.

Yoko Futagi, Paul Deane, Martin Chodorow, and Joel Tetreault. A computational approach to detecting collocation errors in the writing of non-native speakers of English. Computer Assisted Language Learning, 21(4):353–367, 2008.

Jeroen Geertzen, Theodora Alexopoulou, and Anna Korhonen. Automatic linguis- tic annotation of large scale L2 databases: The EF-Cambridge open language database (EFCAMDAT). In Proceedings of the 31st Second Language Research Forum. Somerville, MA: Cascadilla Proceedings Project, 2013.

Marcin Junczys-Dowmunt Roman Grundkiewicz. The AMU system in the CoNLL- 2014 Shared Task: Grammatical error correction by data-intensive and feature- rich statistical machine translation. Proceedings of the 18th CoNLL: Shared Task, page 25, 2014.

Roman Grundkiewicz and Marcin Junczys-Dowmunt. The Wiked Error Corpus: A corpus of corrective Wikipedia edits and its application to grammatical error correction. In International Conference on Natural Language Processing, 2014.

Claudia Leacock, Martin Chodorow, Michael Gamon, and Joel Tetreault. Auto- mated grammatical error detection for language learners. Synthesis Lectures on Human Language Technologies, 7(1):1–170, 2014.

John Lee and Stephanie Seneff. Correcting misuse of verb forms. In Proceedings of ACL-08: HLT, pages 174–182, 2008.
Anne Li-E Liu, David Wible, and Nai-Lung Tsao. Automated suggestions for miscollocations. In Proceedings of the 4th Workshop on Innovative Use of NLP for Building Educational Applications, pages 47–50, 2009.

Tomoya Mizumoto, Yuta Hayashibe, Mamoru Komachi, Masaaki Nagata, and Yuji Matsumoto. The effect of learner corpus size in grammatical error correction of ESL writings. In Proceedings of the 24th Cocling, page 863, 2012.

Diane Nicholls. The Cambridge Learner Corpus: Error coding and analysis for lex- icography and ELT. In Proceedings of the Corpus Linguistics 2003 Conference, volume 16, pages 572–581, 2003.

Alla Rozovskaya and Dan Roth. Algorithm selection and model adaptation for ESL correction tasks. In Proceedings of the 49th ACL, pages 924–933, 2011.

Alla Rozovskaya and Dan Roth. Grammatical error correction: Machine translation and classifiers. Urbana, 51:61820, 2016.

Alla Rozovskaya, Dan Roth, and Vivek Srikumar. Correcting grammatical verb errors. In Proceedings of the 14th EACL, pages 358–367, 2014.

Yu Sawai, Mamoru Komachi, and Yuji Matsumoto. A learner corpus-based approach to verb suggestion for ESL. In Proceedings of the 51st ACL, pages 708–713, 2013.

C-C Shei and Helen Pain. An ESL writer’s collocational aid. Computer Assisted Language Learning, 13(2):167–182, 2000.

Frank Smadja. Retrieving collocations from text: Xtract. Computational linguistics, 19(1):143–177, 1993.

Andreas Stolcke et al. Srilm-an extensible language modeling toolkit. In Inter- speech, volume 2002, page 2002, 2002.

Joel Tetreault, Jennifer Foster, and Martin Chodorow. Using parse features for preposition selection and error detection. In Proceedings of the 48th ACL, pages 353–358, 2010.

Jian-Cheng Wu, Yu-Chia Chang, Teruko Mitamura, and Jason S Chang. Automatic collocation suggestion in academic writing. In Proceedings of the 48th ACL, pages 115–119, 2010.

Jian-Cheng Wu, Tzu-Hsi Yen, Jim Chang, Guan-Cheng Huang, Jim Chang, Hsiang-Ling Hsu, Yu-Wei Chang, and Jason S Chang. NTHU at the CoNLL- 2014 Shared Task. In Proceedings of the 18th CoNLL: Shared Task, pages 91–95, 2014.

Yuanbin Wu and Hwee Tou Ng. Grammatical error correction using integer linear programming. In Proceedings of the 51st ACL, pages 1456–1465, 2013.