Elamur Research⌕
☰
home / papers / 2010 / Aiding Pronoun Translation with Co-Reference Resolution

Aiding Pronoun Translation with Co-Reference Resolution

Summary

Statistical MT systems translate each sentence independently, so they cannot choose the correct gender for English it and they when the antecedent sits in an earlier sentence. This matters for gendered target languages such as French (il vs. elle). Le Nagard and Koehn handle the problem outside the decoder. They run co-reference resolution on the English side, using their own implementations of the Hobbs and Lappin and Leass algorithms, to find each pronoun's antecedent noun. Word alignment, or the translation mapping at test time, identifies the French counterpart of that noun, and the Lefff lexicon supplies its gender. The pronoun is then rewritten as it-masculine, it-feminine or it-neutral (likewise for they). A phrase-based Moses system is trained on the annotated Europarl data and applied to an annotated test set. The annotation moves translation probabilities toward the intended gender, but BLEU and manually checked pronoun accuracy remain essentially at baseline. The main bottleneck is weak co-reference resolution. The study is an early attempt to bring discourse information into SMT.

Key points

  • Data: the Europarl training corpus has 1,393,452 sentences. The test set covers the October 5, 2000 proceedings (1742 sentences), chosen because it keeps whole documents in order.
  • Tools: the Moses toolkit, the Collins parser and the Lefff lexicon (over 80,000 French words with gender and number). Pleonastic it is detected with roughly 83% precision and recall for the referential class.
  • In the baseline, it translates to il with probability 0.307 and to elle with 0.090. With annotation, masculine-marked they prefers ils at about 0.43, and feminine-marked forms now slightly favour feminine French pronouns.
  • In the manual check, only 33 of 59 pronouns received the correct annotation (56%). Correct pronoun translations reached 68%, against 69% for the baseline. Among correctly annotated pronouns, 82% (27/33) were translated correctly.
  • The they-neutral label occurs in 12,424 training sentences. Each of these marks a failed resolution, since French has no neutral plural, which shows how much noise resolution errors add.

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: Coreference resolution

Cited by

65 papers of our library cite it (of the 140673 whose bibliographies we read); 15 of them have a page here, 50 not yet.

Cite

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

@inproceedings{nagard2010aiding,
  title = {{Aiding Pronoun Translation with Co-Reference Resolution}},
  author = {Nagard, Ronan Le and Koehn, Philipp},
  booktitle = {Proceedings of the Joint Fifth Workshop on Statistical Machine Translation and MetricsMATR},
  year = {2010},
  url = {https://aclanthology.org/W10-1737/}
}

References

31 references, 11 of them with a card of ours. ↗ leads to the work, «find» searches for one we have no id for yet.

  1. C. Aone and S.W. Bennett. Evaluating automated and manual acquisition of anaphora resolution strategies. In Proceedings of the 33rd annual meeting on Association for Computational Linguistics, pages 122–129. Association for Computational Linguistics Morristown, NJ, USA, 1995. find our card
  2. S. E. Brennan, M. W. Friedman, and C. J. Pollard. A centering approach to pronouns. In Proceedings of the 25th annual meeting on Association for Computational Linguistics, pages 155–162, 1987. find
  3. E. Charniak. Toward a model of children’s story comprehension. MIT, 1972. find
  4. N. Chinchor and L. Hirschmann. MUC-7 coreference task definition, version 3.0. In Proceedings of MUC, volume 7, 1997. find our card
  5. M. Collins. Head-driven statistical models for natural language parsing. Computational Linguistics, 29(4):589–637, 2003. find our card
  6. N. Ge, J. Hale, and E. Charniak. A statistical approach to anaphora resolution. In Proceedings of the Sixth Workshop on Very Large Corpora, pages 161–170, 1998. find our card
  7. R. Grishman and B. Sundheim. Message understanding conference-6: A brief history. In Proceedings of the 16th conference on Computational Linguistics-Volume 1, pages 466– 471. Association for Computational Linguistics Morristown, NJ, USA, 1996. find
  8. B. J. Grosz, S. Weinstein, and A. K. Joshi. Centering: A framework for modeling the local coherence of discourse. Computational Linguistics, 21(2):203–225, 1995. find our card
  9. J. R. Hobbs. Resolving Pronoun References. Lingua, 44:339–352, 1978. find our card
  10. D. Jurafsky, J. H. Martin, A. Kehler, K. Vander Linden, and N. Ward. Speech and language processing. Prentice Hall New York, 2000. find
  11. A. Kehler, D. Appelt, L. Taylor, and A. Simma. The (non) utility of predicate-argument frequencies for pronoun interpretation. In Proc. of HLT-NAACL, volume 4, pages 289–296, 2004. find
  12. D. Klapholz and A. Lockman. Contextual reference resolution. American Journal of Computational Linguistics, microfiche 36, 1975. find
  13. Philipp Koehn. Europarl: A parallel corpus for statistical machine translation. In Proceedings of the Tenth Machine Translation Summit (MT Summit X), Phuket, Thailand, September 2005. find
  14. Philipp Koehn, Hieu Hoang, Alexandra Birch, Chris Callison-Burch, Marcello Federico, Nicola Bertoldi, Brooke Cowan, Wade Shen, Christine Moran, Richard Zens, Christopher J. Dyer, Ondřej Bojar, Alexandra Constantin, and Evan Herbst. Moses: Open source toolkit for statistical machine translation. In… find
  15. S. Lappin and H.J. Leass. An algorithm for pronominal anaphora resolution. Computational Linguistics, 20(4):561, 1994. find our card
  16. Herbert Leass and Ulrike Schwall. An Anaphora Resolution Procedure for Machine Translation. Technical Report Report 172, IBM Germany Science Center, Institute for Knowledge Based Systems, 1991. find
  17. E. Liddy, S. Bonzi, J. Katzer, and E. Oddy. A study of discourse anaphora in scientific abstracts. Journal of the American Society for Information Science, 38(4):255–261, 1987. find
  18. Michael C. McCord. Slot grammar: A system for simpler construction of practical natural language grammars. In Proceedings of the International Symposium on Natural Language and Logic, pages 118–145, London, UK, 1990. Springer-Verlag. ISBN 3-540-53082-7. find
  19. R. Mitkov, S. K. Choi, and R. Sharp. Anaphora resolution in Machine Translation. In Proceedings of the Sixth International Conference on Theoretical and Methodological Issues in Machine Translation, TMI’95, 1995. find our card
  20. V. Ng. Machine learning for coreference resolution: From local classification to global ranking. In Proceedings of the 43rd Annual Meeting on Association for Computational Linguistics, page 164. Association for Computational Linguistics, 2005. find our card
  21. C. D. Paice and G. D. Husk. Towards the automatic recognition of anaphoric features in English text: the impersonal pronoun. Computer Speech & Language, 2(2):109–132, 1987. find
  22. M. Poesio and R. Artstein. Anaphoric annotation in the ARRAU corpus. In Proceedings of the International Conference on Language Resources and Evaluation (LREC), 2008. find our card
  23. M. Poesio, U. Kruschwitz, and J. Chamberlain. ANAWIKI: Creating anaphorically annotated resources through Web cooperation. In Proceedings of the International Conference on Language Resources and Evaluation (LREC), volume 8. Citeseer, 2008. find
  24. Judita Preiss. Choosing a parser for anaphora resolution. In 4th Discourse Anaphora and Anaphor Resolution Colloquium (DAARC), pages 175– 180. Edi cões Colibri, 2002. find
  25. Long Qiu, Min yen Kan, and Tat seng Chua. A public reference implementation of the rap anaphora resolution algorithm. In Proceedings of the International Conference on Language Resources and Evaluation (LREC), pages 291– 294, 2004. find our card
  26. J. E. Rush, R. Salvador, and A. Zamora. Automatic abstracting and indexing. II. Production of indicative abstracts by application of contextual inference and syntactic coherence criteria. Journal of the American Society for Information Science and Technology, 22(4):260–274, 1971. find
  27. B. Sagot, L. Clément, E. V. de La Clergerie, and P. Boullier. The Lefff 2 syntactic lexicon for French: architecture, acquisition, use. In Proceedings of the International Conference on Language Resources and Evaluation (LREC), 2006. find
  28. Randall Sharp. CAT2 – implementing a formalism for multi-lingual MT. In 2nd International Conference on Theoretical and Methodological Issues in Machine Translation of Natural Language, pages 3–6, 1988. find
  29. M. A. Walker. Centering, anaphora resolution, and discourse structure. Centering theory in discourse, pages 401–435, 1998. find
  30. Y. Wilks. A preferential, pattern-seeking, semantics for natural language inference. Words and Intelligence I, pages 83–102, 1975. find
  31. T. Winograd. Understanding natural language. Cognitive Psychology, 3(1):1–191, 1972. find

↑ back to top