'IMS at the CoNLL 2017 UD Shared Task: CRFs and Perceptrons Meet Neural Networks'

What it is about

This is the IMS system for the CoNLL 2017 raw-text-to-UD task, which finished third overall and second on the surprise languages. Its main claim is that preprocessing decides much of the final score: a CRF tagger for POS and morphology, a neural supertagger, and in some languages their own word and sentence segmentation (a perceptron parser that marks sentence boundaries). The parsing step blends three parsers with different architectures: a perceptron graph-based parser, a perceptron beam transition parser and a neural greedy transition parser, with several instances of each. On the development sets, supertags add 0.88 LAS points on average over 55 treebanks, the CRF tagger adds 0.66, and blending gains two to three points over any single parser, with a search over blend combinations adding 0.49 more.

Ideas

• I1. Supertags as features in en/src/parse.rs: define the Ouchi model-1 tag (dependency label, head direction, presence of left and right dependents), first with gold tags as an upper bound, then with a tagger built from en features; measure EWT dev LAS with and without the tag.
• I2. Blend of three English parsers: the en perceptron arc-standard parser, the gparse model and a second greedy gparse run, combined by weighted arc voting and decoded with Chu-Liu-Edmonds; the parser order is fixed (BTreeMap) so the blend is deterministic; measure EWT test LAS once per release and choose weights by cv on train.
• I3. Multiword-token splits by edit script: for contractions such as "don't" the paper predicts an edit script rather than a fixed table; in en/src/tok.rs count the tokenisation errors on the lowercase ATIS queries (test only) and on EWT dev before deciding whether a learned splitter is worth it.

For Mova

The code is en/src/tok.rs (tokenizer), en/src/ptag.rs (tagger), en/src/parse.rs (parser), and gparse/src/eisner.rs (decoder). The chapters are 20. Syntax: the Universal Dependencies frame and 07. Punctuation, capital letters, abbreviations and numbers. Measure first the supertag gain with gold tags on EWT dev, since it bounds what any supertag tagger can add.

Topics: Universal Dependencies · Part-of-speech tagging

Cite

The paper itself, not this card. BibTeX from the library's record of the work. DOI 10.18653/v1/k17-3004

@inproceedings{bjorkelund2017ims,
  title = {{'IMS at the CoNLL 2017 UD Shared Task: CRFs and Perceptrons Meet Neural Networks'}},
  author = {Björkelund, Anders and Falenska, Agnieszka and Yu, Xiang and Kuhn, Jonas},
  booktitle = {Proceedings of the CoNLL 2017 Shared Task: Multilingual Parsing from Raw Text to Universal Dependencies},
  year = {2017},
  doi = {10.18653/v1/k17-3004},
  url = {https://doi.org/10.18653/v1/k17-3004}
}

References

28 references, 6 of them with a card of ours. ↗ leads to the work, «find» searches for one we have no id for yet.

1. A. Martins, M. Almeida, and N. A. Smith. 2013. ”turning on the turbo: Fast third-order non-projective turbo parsers”. In Annual Meeting of the Association for Computational Linguistics - ACL. volume -, pages 617–622. find our card
2. Anders Björkelund and Joakim Nivre. 2015. Non-deterministic oracles for unrestricted non-projective transition-based dependency parsing. In Proceedings of the 14th International Conference on Parsing Technologies. Association for Computational Linguistics, Bilbao, Spain, pages 76–86. http://www.acl… ↗
3. Ryan McDonald and Fernando Pereira. 2006. Online learning of approximate dependency parsing algorithms. In Proceedings of the 11th Conference of the European Chapter of the ACL (EACL 2006). Association for Computational Linguistics, Trento, Italy, pages 81–88. http://www.aclweb.org/anthology-new/E/E… ↗ our card
4. Bernd Bohnet. 2010. Top accuracy and fast dependency parsing is not a contradiction. In Proceedings of the 23rd International Conference on Computational Linguistics (Coling 2010). find
5. Coling 2010 Organizing Committee, Beijing, China, pages 89–97. http://www.aclweb.org/anthology/C10-1011. ↗
6. Thomas Müller, Helmut Schmid, and Hinrich Schütze. 2013. Efficient Higher-Order CRFs for Morphological Tagging. In In Proceedings of EMNLP. find
7. Xavier Carreras. 2007. Experiments with a higher-order projective dependency parser. In Proceedings of the CoNLL Shared Task Session of EMNLP-CoNLL 2007. Association for Computational Linguistics, Prague, Czech Republic, pages 957– 961. http://www.aclweb.org/anthology/D/D07/D07-1101. ↗
8. Joakim Nivre. 2009. Non-projective dependency parsing in expected linear time. In Proceedings of the Joint Conference of the 47th Annual Meeting of the ACL and the 4th International Joint Conference on Natural Language Processing of the AFNLP. Association for Computational Linguistics, Suntec, Singa… find
9. Yoeng-jin Chu and Tseng-hong Liu. 1965. On the shortest aborescence of a directed graph. Science Sinica 14:1396–1400. find
10. Joakim Nivre, Željko Agić, Lars Ahrenberg, et al. 2017. Universal dependencies 2.0 CoNLL 2017 shared task development and test data. LIN-DAT/CLARIN digital library at the Institute of Formal and Applied Linguistics, Charles University. http://hdl.handle.net/11234/1-2184. ↗
11. Jack Edmonds. 1967. Optimum branchings. Journal of Research of the National Bureau of Standards 71(B):233–240. find
12. Jason Eisner. 1997. Bilexical grammars and a cubic-time probabilistic parser. In Proceedings of the 5th International Workshop on Parsing Technologies (IWPT). MIT, Cambridge, MA, pages 54–65. http://cs.jhu.edu/ jason/papers/\#eisner-1997-iwpt. ↗
13. Joakim Nivre, Marco Kuhlmann, and Johan Hall. 2009. An improved oracle for dependency parsing with online reordering. In Proceedings of the 11th International Conference on Parsing Technologies (IWPT’09). Association for Computational Linguistics, Paris, France, pages 73–76. http://www.aclweb.org/an… ↗
14. Agnieszka Faleńska, Anders Björkelund, Özlem Çetinoğlu, and Wolfgang Seeker. 2015. Stacking or supertagging for dependency parsing – what’s the difference? In Proceedings of the 14th International Conference on Parsing Technologies. Association for Computational Linguistics, Bilbao, Spain, page… find
15. Hiroki Ouchi, Kevin Duh, and Yuji Matsumoto. 2014. Improving Dependency Parsers with Supertags. In Proceedings of the 14th Conference of the European Chapter of the Association for Computational Linguistics, volume 2: Short Papers. Association for Computational Linguistics, Gothenburg, Sweden, pages… find
16. Liang Huang, Suphan Fayong, and Yang Guo. 2012. Structured perceptron with inexact search. In Proceedings of the 2012 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies. Association for Computational Linguistics, Montréal, Canada,… find our card
17. Martin Potthast, Tim Gollub, Francisco Rangel, Paolo Rosso, Efstathios Stamatatos, and Benno Stein. 2014. Improving the reproducibility of PAN’s shared tasks: Plagiarism detection, author identification, and author profiling. In Evangelos Kanoulas, Mihai Lupu, Paul Clough, Mark Sander-son, Mark Hall… — [doi:10.1007/978-3-319-11382-1](https://doi.org/10.1007/978-3-319-11382-1) find
18. Marie-Catherine de Marneffe, Manuela Sanguinetti, Maria Simi, Hiroshi Kanayama, Valeria de Paiva, Kira Droganova, Hěctor Martı́nez Alonso, Hans Uszkoreit, Vivien Macketanz, Aljoscha Burchardt, Kim Harris, Katrin Marheinecke, Georg Rehm, Tolga Kayadelen, Mohammed Attia, Ali Elkahky, Zhuoran Yu, Emil… find
19. CoNLL 2017 Shared Task: Multilingual Parsing from Raw Text to Universal Dependencies. In Proceedings of the CoNLL 2017 Shared Task: Multilingual Parsing from Raw Text to Universal Dependencies. Association for Computational Linguistics. find our card
20. Rudolf Rosa and Zdenek Zabokrtský. 2015. Klcpos3 - a language similarity measure for delexicalized parser transfer. In Proceedings of the 53rd Annual Meeting of the Association for Computational Linguistics and the 7th International Joint Conference on Natural Language Processing of the Asian Feder… find
21. Kenji Sagae and Alon Lavie. 2006. Parser combination by reparsing. In Proceedings of the Human Language Technology Conference of the NAACL, Companion Volume: Short Papers. Association for Computational Linguistics, New York City, USA, pages 129–132. http://www.aclweb.org/anthology/N/N06/N06-2033. ↗
22. Daniel Zeman and Philip Resnik. 2008. Cross-language parser adaptation between related languages. In Third International Joint Conference on Natural Language Processing, IJCNLP 2008, Hyderabad, India, January 7-12, 2008. pages 35–42. http://aclweb.org/anthology/I/I08/I08-3008.pdf. ↗
23. Milan Straka, Jan Hajič, and Jana Straková. 2016. UDPipe: trainable pipeline for processing CoNLL-U files performing tokenization, morphological analysis, pos tagging and parsing. In Proceedings of the Tenth International Conference on Language Resources and Evaluation (LREC’16). European Language… find our card
24. Yuan Zhang and David Weiss. 2016. Stack-propagation: Improved representation learning for syntax. In Proceedings of the 54th Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers). Association for Computational Linguistics, pages 1557–1566. https://doi.org/10.18653/v… find
25. Yue Zhang and Stephen Clark. 2008. A tale of two parsers: Investigating and combining graph-based and transition-based dependency parsing. In Proceedings of the 2008 Conference on Empirical Methods in Natural Language Processing. Association for Computational Linguistics, Honolulu, Hawaii, pages 562… find our card
26. Mihai Surdeanu and Christopher D Manning. 2010. Ensemble models for dependency parsing: cheap and good? In Human Language Technologies: The 2010 Annual Conference of the North American Chapter of the Association for Computational Linguistics. Association for Computational Linguistics, pages 649–652. find
27. Xiang Yu, Agnieszka Falenska, and Ngoc Thang Vu. 2017. A general-purpose tagger with convolutional neural networks. In arXiv preprint arXiv:1706.01723. find
28. Xiang Yu and Ngoc Thang Vu. 2017. Character composition model with convolutional neural networks for dependency parsing on morphologically rich languages. In Proceedings of the 52nd Annual Meeting of the Association for Computational Linguistics. Association for Computational Linguistics, Vancouver,… find

This page in Markdown → https://md.mova.elamur.ai/free/en/docs/papers/ud-2017-bjorkelund-ims-conll-2017-ud-shared/