Elamur Research⌕
☰
home / papers / 2022 / A Learning-Based Dependency to Constituency Conversion Algorithm for the Turkish Language

A Learning-Based Dependency to Constituency Conversion Algorithm for the Turkish Language

Summary

The authors propose the first algorithm that converts dependency trees into constituency trees and is tailored specifically to Turkish, a language with few syntactic resources. The input is the manually annotated treebank UD Turkish-Penn. It consists of Penn Treebank sentences that professional translators translated into Turkish and that linguists annotated according to the UD guidelines. The algorithm builds the tree bottom-up. First, a subtree is created for each dependency edge. The subtrees are then combined through three actions: attach the left neighbor (Left), attach the right neighbor (Right), or merge nodes into a phrasal projection (Merge). An oracle decides which action to perform. The oracle is either a set of predefined heuristics or a feature-based machine learning model enhanced with bagging. The features are drawn from dependency types, headedness, POS and morphological analysis. The trees are deliberately flat, and phrase tags are determined from morphology. Quality was evaluated against gold-standard trees manually annotated by a team of linguists. The result is more than 8000 constituency trees parallel to the UD annotation. The work establishes a baseline for Turkish. It also demonstrates a simple way to obtain training data for constituency parsers in low-resource languages.

Key points

  • The input corpus is UD Turkish-Penn. The authors deliberately chose manual annotation rather than automatically generated dependencies. In automatic annotation, the head of contentious constructions such as "genel olarak" is often assigned inconsistently.
  • Morphological features and POS are provided by the open-source Turkish morphological analyzer of Yıldız et al. (2019). The bootstrap aggregating (bagging) meta-algorithm is added to the conversion model.
  • The trees are flatter than in X-Bar or the Penn Treebank. Ternary branching is allowed, and there are no X' levels or empty projections. Complements and adjuncts of the same head are placed at the same level, following the strategy of Hajič et al. (1998).
  • Phrases carry only basic tags such as NP, VP, PP, DP. There are no function tags like NP-SUBJ or ADV-TMP, since the predicate-argument structure is meant to be described by the Turkish PropBank (TRopBank).
  • The more than 8000 resulting constituency trees are released on GitHub together with the corresponding UD dependency trees.

Our summary of the paper, in our own words — not its text. The work belongs to its authors; read it at the links above.

Fields: Linguistics & languages

Topics: Treebanks

Cite

The paper itself, not this page. BibTeX from the library's record of the work.

@inproceedings{marsan2022learning,
  title = {{A Learning-Based Dependency to Constituency Conversion Algorithm for the Turkish Language}},
  author = {Marşan, Büşra and Yıldız, Oğuz K. and Kuzgun, Aslı and Cesur, Neslihan and Yenice, Arife B. and Sanıyar, Ezgi and Kuyrukçu, Oğuzhan and Arıcan, Bilge N. and Yıldız, Olcay Taner},
  booktitle = {Proceedings of the Thirteenth Language Resources and Evaluation Conference},
  year = {2022},
  doi = {10.63317/5cngxuqy3u9d},
  url = {https://doi.org/10.63317/5cngxuqy3u9d}
}

References

27 references, 3 of them with a card of ours. ↗ leads to the work, «find» searches for one we have no id for yet.

  1. Bai, J., Wang, Y., Chen, Y., Yang, Y., Bai, J., Yu, J., and Tong, Y. (2021). Syntax-bert: Improving pretrained transformers with syntax trees. arXiv. find
  2. Bick, E. (2006). Turning a dependency treebank into a PSG-style constituent treebank. In Proceedings of the Fifth International Conference on Language Resources and Evaluation (LREC’06), Genoa, Italy, May. European Language Resources Association (ELRA). find
  3. Black, E., Abney, S., Flickenger, D., Gdaniec, C., Grishman, R., Harrison, P., Hindle, D., Ingria, R., Jelinek, F., Klavans, J., Liberman, M., Marcus, M., Roukos, S., Santorini, B., and Strzalkowski, T. (1991). A procedure for quantitatively comparing the syntactic coverage of English grammars. In S… find
  4. Bosco, C. (2007). Multiple-step treebank conversion: From dependency to Penn format. In Proceedings of the Linguistic Annotation Workshop, pages 164– 167, Prague, Czech Republic, June. Association for Computational Linguistics. find
  5. Bouma, G. and Kloosterman, G. (2002). Querying dependency treebanks in XML. In Proceedings of the Third International Conference on Language Resources and Evaluation (LREC’02), Las Palmas, Canary Islands - Spain, May. European Language Resources Association (ELRA). find
  6. Cahill, A., Burke, M., O’Donovan, R., Riezler, S., van Genabith, J., and Way, A. (2008). Wide-coverage deep statistical parsing using automatic dependency structure annotation. Computational Linguistics, 34(1):81–124. Candito, M., Crabbé, B., and Denis, P. (2010). Statistical French dependency pars… find
  7. Chomsky, N. (1957). Syntactic structures (the hague: Mouton, 1957). Review of Verbal Behavior by BF Skinner, Language, 35:26–58. find
  8. Chomsky, N. (1968). Remarks on Nominalization. find
  9. Linguistics Club, Indiana University. Daum, M., Foth, K., and Menzel, W. (2004). Automatic transformation of phrase treebanks to dependency trees. 06. find
  10. Ding, Y. and Palmer, M. (2004). Synchronous dependency insertion grammars: A grammar formalism for syntax based statistical MT. In Proceedings of the Workshop on Recent Advances in Dependency Grammar, pages 90–97, Geneva, Switzerland, aug 28. find
  11. COLING. Ding, Y. and Palmer, M. (2005). Machine translation using probabilistic synchronous dependency insertion grammars. In Proceedings of the 43rd Annual Meeting of the Association for Computational Linguistics (ACL’05), pages 541–548, Ann Arbor, Michigan, June. Association for Computational Ling… find our card
  12. Grammar, C. and Hockenmaier, J. (2003). Data and models for statistical parsing with combinatory categorial grammar. 09. Hajic, J., Brill, E., Collins, M., Hladká, B., Jones, D., Kuo, C., Ramshaw, L., Schwartz, O., Tillmann, C., and Zeman, D. (1998). Core natural language processing technology appl… find
  13. Huang, L., Knight, K., and Joshi, A. (2006). Statistical syntax-directed translation with extended domain of locality. 01. Kara, N., Aslan, D. B., Marşan, B., Bakay, Ö., Ak, K., and Yıldız, O. T. (2020). TRopBank: Turkish Prop-Bank v2.0. In Proceedings of the 12th Language Resources and Evaluation… find our card
  14. European Language Resources Association. Koehn, P., Och, F. J., and Marcu, D. (2003). Statistical phrase-based translation. find
  15. University Of Southern California Marina Del Rey Information Sciences Inst. Kong, L., Rush, A. M., and Smith, N. A. (2015). Transforming dependencies into phrase structures. In Proceedings of the 2015 Conference of the North American Chapter of the Association for Computational Linguistics: Human La… find
  16. Lee, Y.-S. and Wang, Z. (2016). Language independent dependency to constituent tree conversion. In COLING. find
  17. Lu, A., Malamud, S., and Xue, N. (2016). Converting syntagrus dependency treebank into penn treebank style. pages 16–21, 01. find
  18. Matthews, P. H. (1981). Syntax. cambridge etc. Cambridge University Press (Cambridge Textbooks in Linguistics). Indices. Nivre, J., de Marneffe, M.-C., Ginter, F., Hajič, J., Manning, C. D., Pyysalo, S., Schuster, S., Tyers, F., and Zeman, D. (2020). Universal dependencies v2: An evergrowing multil… — [arXiv:2004.10643](https://arxiv.org/abs/2004.10643) find
  19. Nivre, J. (2003). Theory-supporting treebanks. In Proceedings of the 2nd Workshop on Treebanks and Linguistic Theories, pages 117–128. find
  20. Percival, W. K. (1990). Reflections on the history of dependency notions in linguistics. Historiographia linguistica, 17(1-2):29–47. find
  21. Sag, I., Wasow, T., and Bender, E. (2003). Syntactic theory: A formal introduction, 2nd edition. Bib-liovault OAI Repository, the University of Chicago Press, 01. find
  22. Sampson, G. and Babarczy, A. (2002). A test of the leaf-ancestor metric for parse accuracy. In The Workshop Programme, page 23. find
  23. Shen, L., Xu, J., and Weischedel, R. (2008). A new string-to-dependency machine translation algorithm with a target dependency language model. In Proceedings of ACL-08: HLT, pages 577–585, Columbus, Ohio, June. Association for Computational Linguistics. find our card
  24. Wang, W. and Harper, M. P. (2004). A statistical constraint dependency grammar (cdg) parser. In Proceedings of the Workshop on Incremental Parsing: Bringing Engineering and Cognition Together, In-crementParsing ’04, pages 42–49, USA. Association for Computational Linguistics. find
  25. Xia, F. and Palmer, M. (2001). Converting dependency structures to phrase structures. In Proceedings of the First International Conference on Human Language Technology Research. find
  26. Xie, J., Mi, H., and Liu, Q. (2011). A novel dependency-to-string model for statistical machine translation. In Proceedings of the 2011 Conference on Empirical Methods in Natural Language Processing, pages 216–226, Edinburgh, Scotland, UK., July. Association for Computational Linguistics. find
  27. Yıldız, O. T., Avar, B., and Ercan, G. (2019). An open, extendible, and fast Turkish morphological analyzer. In Proceedings of the International Conference on Recent Advances in Natural Language Processing (RANLP 2019), pages 1364–1372, Varna, Bulgaria, September. INCOMA Ltd. find

↑ back to top

This page in Markdown → https://md.research.elamur.ai/free/en/papers/an-2022-learning-based-dependency-constituency-conversion/