我們提出了一種解釋英文句子校正原因的方法，目標是根據錯誤類型、問題詞和上下文客製化校正的解釋。在我們的方法中，我們會分析經過校正的句子並且偵測問題類型和問題詞。方法的主要步驟包含：分析錯誤類型和問題詞、產生各種錯誤類型的解釋樣板和找到錯誤對應的文法、搭配詞與例句用以套入解釋樣板。在執行時，我們首先對帶有校正標記的句子進行錯誤類型分類，然後針對該錯誤找到對應的解釋模板並提供詳盡的解釋。我們做出了一個解釋校正的雛型系統，TellMeWhy，可以應用於帶有校正標記的語料庫或者與文法改錯系統結合，先將含有錯誤的句子送進文法改錯系統校正，再將校正過後的文句交給TellMeWhy解釋校正原因。我們使用常見錯誤類型的句子當作測試資料，初步的評估發現此方法有潛力改進現有的商業寫作服務系統。

We introduce a method for learning to provide feedback on a given English sentence with a correction. In our approach, the sentence with a correction is analyzed to identify the error type and a problem word, aimed at customizing explanations according to the context of the error type and the problem word. The method involves learning the relation of errors and problem words, generating common feedback patterns of each error type, and extracting grammar patterns along with collocations and example sentences for explanation.
At run-time, a sentence with a correction is classified, and the problem word and template are identified to provide detailed explanations. We present a prototype feedback system, TellMeWhy, that applies the method to an annotated corpus and an underlying grammatical error correction system. Preliminary evaluation on a set of representative sentences with common errors shows that the method has potential to improve existing commercial writing services.

Abstract------------------------------------i
摘要----------------------------------------ii
致謝----------------------------------------iii
Contents-----------------------------------v
List of Figures------------------------------vii
List of Tables-------------------------------viii
1 Introduction-------------------------------1
2 Related Work-----------------------------7
3 Methodology-----------------------------11
4 Experiment and Evaluation----------------17
5 Future Work and Summary----------------23
Appendices--------------------------------25
Reference----------------------------------29

1. Øistein E Andersen, Helen Yannakoudakis, Fiona Barker, and Tim Parish. Devel-oping and testing a self-assessment and tutoring system. InProceedings of theeighth workshop on innovative use of NLP for building educational applications,pages 32–41, 2013
2. John Bitchener and Dana R Ferris.Written corrective feedback in second languageacquisition and writing. Routledge, 2012
3. John Bitchener, Stuart Young, and Denise Cameron. The effect of different typesof corrective feedback on esl student writing.Journal of second language writing,14(3):191–205, 2005
4. Christopher Bryant, Mariano Felice, and Ted Briscoe. Automatic annotation andevaluation of error types for grammatical error correction. InProceedings of the55th Annual Meeting of the Association for Computational Linguistics (Volume1: Long Papers), volume 1, pages 793–805, 2017
5. Jim Chang and Jason Chang. Writeahead2: Mining lexical grammar patterns forassisted writing. InProceedings of the 2015 Conference of the North Ameri-can Chapter of the Association for Computational Linguistics: Demonstrations,pages 106–110, 2015
6. Jhih-Jie Chen, Jim Chang, Ching-Yu Yang, Mei-Hua Chen, and Jason S Chang.Extracting formulaic expressions and grammar and edit patterns to assist aca-demic writing.EUROPHRAS 2017: Computational and Corpus-based Phrase-ology: Recent Advances and Interdisciplinary Approaches, 2017
7. Collins CoBUILD. Collins cobuild advanced learner’s dictionary.Aufl. Glasgow,2006
8. Collins Cobuild et al.Collins Cobuild English Grammar. Collins Cobuild, 2005
9. Nina Garrett. A psycholinguistic perspective on grammar and call.Modern Me-dia in Foreign Language Education: Theory and Implementation. The ACTFLForeign Language Education Series., 1982
10. Jeroen Geertzen, Theodora Alexopoulou, and Anna Korhonen. Automatic lin-guistic annotation of large scale l2 databases: The ef-cambridge open languagedatabase (efcamdat). InProceedings of the 31st Second Language Research Fo-rum. Somerville, MA: Cascadilla Proceedings Project, 2013
11. Yan Huang, Akira Murakami, Theodora Alexopoulou, and Anna Korhonen. De-pendency parsing of learner english.International Journal of Corpus Linguistics,23(1):28–54, 2018
12. Francis Gill Hunston, Susan and Elizabeth Manning.Collins COBUILD GrammarPatterns 1: Verbs. HarperCollins, 1996
13. Francis Gill Hunston, Susan and Elizabeth Manning.Collins COBUILD GrammarPatterns 2: Nouns and Adjetives. HarperCollins, 1997
14. Susan Hunston and Gill Francis.Pattern grammar: A corpus-driven approach tothe lexical grammar of English, volume 4. John Benjamins Publishing, 2000
15. Claudia Leacock, Martin Chodorow, Michael Gamon, and Joel Tetreault. Auto-mated grammatical error detection for language learners.Synthesis lectures onhuman language technologies, 3(1):1–134, 2010
16. Noriko Nagata. Intelligent computer feedback for second language instruction.The modern language journal, 77(3):330–339, 1993
17. Diane Nicholls. The cambridge learner corpus: Error coding and analysis forlexicography and elt. InProceedings of the Corpus Linguistics 2003 conference,volume 16, pages 572–581, 2003
18. Frank Smadja. Retrieving collocations from text: Xtract.Computational linguis-tics, 19(1):143–177, 1993
19. Yukio Tono, Yoshiho Satake, and Aika Miura. The effects of using corpora onrevision tasks in l2 writing with coded error feedback.ReCALL, 26(2):147–162,2014
20. John Truscott. The effect of error correction on learners’ ability to write accurately.Journal of second language Writing, 16(4):255–272, 2007
21. ND Turton and JB Heaton. Longman dictionary of common errors. new edition,1996
22. Helen Yannakoudakis, Ted Briscoe, and Ben Medlock.A new dataset andmethod for automatically grading esol texts. InProceedings of the 49th Annual Meeting of the Association for Computational Linguistics: Human Language Technologies-Volume 1, pages 180–189. Association for ComputationalLinguistics, 2011
23. Tzu-Hsi Yen, Jian-Cheng Wu, Jim Chang, Joanne Boisson, and Jason Chang.Writeahead: Mining grammar patterns in corpora for assisted writing.Proceed-ings of ACL-IJCNLP 2015 System Demonstrations, pages 139–144, 2015