Elamur Research⌕
☰
home / papers / 2014 / Inducing Discourse Connectives from Parallel Texts

Inducing Discourse Connectives from Parallel Texts

Summary

Lists of discourse connectives (DCs), such as however or because, help both discourse theory and discourse parsing, but few languages have them. Laali and Kosseim propose building such a list automatically from parallel text. They assume that discourse relations survive translation. First, they ran the PDTB-style End-To-End parser of Lin et al. (2010) on the English side of English–French Europarl. They kept only sentence pairs whose English half had exactly one explicit relation, and that relation then labels the French sentence. Next, every French n-gram of length one to six becomes a candidate. Candidates are ranked by Log-Likelihood Ratio (LLR) association with the relation label, an idea borrowed from collocation extraction. A frequency threshold and one of three filters (word alignment, POS patterns or syntax trees) then prune the list. Against the LEXCONN lexicon, the syntax-based filter clearly performs best, reaching a MAP of 0.39. It also surfaces valid connectives that LEXCONN lacks. The method could therefore help bootstrap DC resources for languages with parallel data but no manual lexicon.

Key points

  • Corpus: Filtering Europarl from 2,054K to 543K sentence pairs (about 14M English and 15M French words) still produced an automatically annotated corpus roughly 30 times larger than PDTB. The parser labels explicit DCs with 81.19% precision on PDTB Section 23.
  • Ranking: LLR was computed with the NSP package. The Frequency filter (threshold 10) leaves only 66% of LEXCONN DCs usable, and 14% never occur in the corpus at all.
  • MAP with Exact Match:
    • LLR alone: 0.06
    • Word-Alignment filter: 0.10
    • POS Pattern filter: 0.12
    • Syntax Tree filter: 0.39 (0.44 when connective fragments are excluded from evaluation)
  • Main error source: Longer candidates often combine a true connective with a stray word (e.g. "mais je"). Only syntactic constituent information reliably removes these.
  • Coverage beyond LEXCONN: Among the top 100 candidates that LEXCONN marked as wrong, 31 were genuine French DCs, such as "toutefois", "certes" and "au lieu de cela".

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: Discourse analysis

Cited by

8 papers of our library cite it (of the 140673 whose bibliographies we read); 3 of them have a page here, 5 not yet.

Cite

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

@inproceedings{laali2014inducing,
  title = {{Inducing Discourse Connectives from Parallel Texts}},
  author = {Laali, Majid and Kosseim, Leila},
  booktitle = {Proceedings of COLING 2014, the 25th International Conference on Computational Linguistics: Technical Papers},
  year = {2014},
  url = {https://aclanthology.org/C14-1058/}
}

References

33 references, 10 of them with a card of ours. ↗ leads to the work, «find» searches for one we have no id for yet.

  1. Amal Al-Saif and Katja Markert. 2010. The Leeds Arabic Discourse Treebank: Annotating Discourse Connectives for Arabic. In LREC, pages 2046–2053, Valletta, Malta. find our card
  2. Laura Alonso Alemany, Irene Castellón Masalles, and Lluı̀s Padró Cirera. 2002. Lexicón computacional de marcadores del discurso. Procesamiento del Lenguaje Natural, 29. find
  3. Bruno Cartoni. 2013. Annotating the meaning of discourse connectives by looking at their translation: The translation-spotting technique. Dialogue & Discourse, 4(2):65–86. find our card
  4. Stefan Evert. 2004. The Statistics of Word Cooccurrences: Word Pairs and Collocations. PhD dissertation, Institut fr Maschinelle Sprachverarbeitung, University of Stuttgart. find
  5. Qin Gao and Stephan Vogel. 2008. Parallel Implementations of Word Alignment Tool. In Software Engineering, Testing, and Quality Assurance for Natural Language Processing, pages 49–57, Columbus, OH, USA. find
  6. Spence Green, Marie-Catherine de Marneffe, John Bauer, and Christopher D. Manning. 2011. Multiword Expression Identification with Tree Substitution Grammars: A Parsing tour de force with French. In Proceedings of the Conference on Empirical Methods in Natural Language Processing, pages 725–735, Edin… find
  7. Adam Kilgarriff, Vojtch Kov, Simon Krek, Irena Srdanovi, and Carole Tiberius. 2010. A Quantitative Evaluation of Word Sketches. In Proceedings of the 14th EURALEX International Congress, Leeuwarden, The Netherlands. find
  8. Alistair Knott. 1996. A data-driven methodology for motivating a set of coherence relations. PhD dissertation, University of Edinburgh. find
  9. Philipp Koehn, Hieu Hoang, Alexandra Birch, Chris Callison-Burch, Marcello Federico, Nicola Bertoldi, Brooke Cowan, Wade Shen, Christine Moran, Richard Zens, Chris Dyer, Ondrej Bojar, Alexandra Constantin, and Evan Herbst. 2007. Moses: Open source toolkit for statistical machine translation. In Proc… find
  10. Philipp Koehn. 2005. Europarl: A parallel corpus for statistical machine translation. In MT summit, volume 5, pages 79–86. find
  11. Ziheng Lin, Hwee Tou Ng, and Min-Yen Kan. 2010. A PDTB-styled end-to-end discourse parser. Natural Language Engineering, 20(02):151–184. find
  12. Christopher D. Manning, Prabhakar Raghavan, and Hinrich Schütze. 2008. Introduction to Information Retrieval, volume 1. Cambridge University Press. find
  13. Thomas Meyer and Bonnie Webber. 2013. Implicitation of Discourse Connectives in (machine) Translation. In Proceedings of the 1st DiscoMT Workshop at ACL 2013 (51st Annual Meeting of the Association for Computational Linguistics), pages 19–26, Sofia, Bulgaria. find our card
  14. Thomas Meyer, Charlotte Roze, Bruno Cartoni, L. Danlos, and A. Popescu-Belis. 2011. Disambiguating discourse connectives using parallel corpora: senses vs. translations. In Proceedings of Corpus Linguistics. find
  15. Thomas Meyer. 2011. Disambiguating Temporal-Contrastive Discourse Connectives for Machine Translation. In Proceedings of ACL-HLT, pages 46–51, Portland, OR, USA. find
  16. Lucie Mladová, Sarka Zikanova, and Eva Hajicová. 2008. From Sentence to Discourse: Building an Annotation Scheme for Discourse Based on Prague Dependency Treebank. In Proceedings of the Sixth International Conference on Language Resources and Evaluation (LREC’08), pages 28–30, Morocco, Marrakech. find
  17. F.J. Och and H. Ney. 2003. A systematic comparison of various statistical alignment models. Computational linguistics, 29(1):19–51. find
  18. Umangi Oza, Rashmi Prasad, Sudheer Kolachina, Dipti Misra Sharma, and Aravind Joshi. 2009. The Hindi Discourse Relation Bank. In Proceedings of the Third Linguistic Annotation Workshop, pages 158–161, Suntec, Singapore. find our card
  19. P. Pecina. 2010. Lexical association measures and collocation extraction. Language Resources and Evaluation, 44(1):137–158. find
  20. T. Pedersen, S. Banerjee, B. T. McInnes, S. Kohli, M. Joshi, and Y. Liu. 2011. The Ngram Statistics Package (text:: NSP)-A Flexible Tool for Identifying Ngrams, Collocations, and Word Associations. In Workshop on Multiword Expression: from Parsing and Generation to the Real World (MWE 2011), pages 1… find
  21. Emily Pitler and Ani Nenkova. 2009. Using syntax to disambiguate explicit discourse connectives in text. In Proceedings of the ACL-IJCNLP 2009 Conference Short Papers , pages, pages 13–16, Suntec, Singapore. find our card
  22. Andrei Popescu-Belis, Thomas Meyer, Jeevanthi Liyanapathirana, Bruno Cartoni, and Sandrine Zufferey. 2012. Discourse-level Annotation over Europarl for Machine Translation: Connectives and Pronouns. In Proceedings of the Eight International Conference on Language Resources and Evaluation (LREC’12),… find
  23. Rashmi Prasad, Nikhil Dinesh, Alan Lee, Eleni Miltsakaki, Livio Robaldo, Aravind K. Joshi, and Bonnie L. Webber. 2008. The Penn Discourse TreeBank 2.0. In Proceedings of the Sixth International Conference on Language Resources and Evaluation (LREC’08), pages 28–30, Marrakech, Morocco. find our card
  24. Rashmi Prasad, Aravind Joshi, and Bonnie Webber. 2010. Realization of Discourse Relations by Other Means: Alternative Lexicalizations. In COLING ’10, pages 1023–1031, Beijing, China. find our card
  25. Charlotte Roze, Laurence Danlos, and Philippe Muller. 2012. LEXCONN: a French lexicon of discourse connectives. Discours [En ligne], (10). find our card
  26. V. Seretan. 2010. Syntax-Based Collocation Extraction, volume 44. Springer-Verlag. find
  27. Manfred Stede and Carla Umbach. 1998. DiMLex: A lexicon of discourse markers for text generation and understanding. In Proceeding of the 17th international conference on Computational Linguistics (COLING-98), pages 1238–1242, Montreal, Canada. Association for Computational Linguistics. find our card
  28. Kristina Toutanova, Dan Klein, Christopher D. Manning, and Yoram Singer. 2003. Feature-rich Part-of-speech Tagging with a Cyclic Dependency Network. In Proceedings of HLT-NAACL 2003, pages 173–180, Edmonton. Association for Computational Linguistics. find
  29. Yannick Versley. 2010. Discovery of ambiguous and unambiguous discourse connectives via annotation projection. In Proceedings of Workshop on Annotation and Exploitation of Parallel Corpora (AEPC), pages 83–82, Tartu, Estonia. Northern European Association for Language Technology (NEALT). find
  30. Yannick Versley. 2011. Towards Finer-grained Tagging of Discourse Connectives. In Beyond Semantics: Corpus-based investigations of pragmatic and discourse phenomena. find
  31. Deniz Zeyrek, In Demirahin, Ay Sevdik-all, Hale gel Balaban, hsan Yalnkaya, and mit Deniz Turan. 2010. The annotation scheme of the Turkish Discourse Bank and an evaluation of inconsistent annotations. In Proceedings of the Fourth Linguistic Annotation Workshop, pages 282–289, Uppsala, Sweden. Assoc… find
  32. Yuping Zhou and Nianwen Xue. 2012. PDTB-style Discourse Annotation of Chinese Text. In Proceedings of the 50th Annual Meeting of the Association for Computational Linguistics: Long Papers-Volume 1, pages 69–77, Jeju, Republic of Korea. Association for Computational Linguistics. find
  33. Lanjun Zhou, Wei Gao, Binyang Li, Zhongyu Wei, and Kam-Fai Wong. 2012. Cross-lingual identification of ambiguous discourse connectives for resource-poor language. In Proceedings of COLING 2012. find our card

↑ back to top

This page in Markdown → https://md.research.elamur.ai/free/papers/laali-2014-inducing-discourse-connectives-parallel-texts/