Elamur Research
☰
home / papers / 2019 / JW300: A Wide-Coverage Parallel Corpus for Low-Resource Languages

JW300: A Wide-Coverage Parallel Corpus for Low-Resource Languages

Written for the Research library (claude-opus-5-5) from the paper text. Original — aclanthology.org/P19-1310.pdf

Summary

Cross-lingual NLP for low-resource languages depends on parallel text, and the scarcity of such text hampers both training and evaluation. As a result, methods aimed at low-resource languages are often tested only on well-resourced ones. Agić and Vulić address this with JW300, a corpus crawled from every publication on jw.org, mostly the Watchtower and Awake! magazines. These articles are largely translated from English and share article IDs across languages. The texts were split and tokenized with Polyglot, using a related language's model when no model existed for a language. Sentences within matching articles were aligned with Yasa.

The authors tested the corpus in two ways:

  • Bilingual lexicon induction: seed dictionaries taken from JW300 let a simple supervised mapping beat a strong unsupervised method.
  • Part-of-speech tagging: basic multi-source annotation projection clearly outperformed Bible-based projection and nearly matched the more complex DsDs system.

The corpus balances many languages with substantial data per pair, though it carries a known religious bias. This makes it a practical resource for multilingual research and for studying bias in NLP.

Key points

  • Size: JW300 covers 343 languages, with about 1.34M articles, roughly 109M sentences and 1.48B tokens. It contains 54,376 aligned language pairs, averaging about 92K parallel sentences per pair.
  • Licensing: the corpus is free for non-commercial use under the copyright holder's terms, which are close to CC-BY-NC-SA.
  • Lexicon induction setup: the BLI experiments used EN, ET, HR, MR and MT, fastText embeddings, and IBM1-derived seed dictionaries.
  • Lexicon induction results: the supervised method (Smith et al., 2017) beat the unsupervised one (Artetxe et al., 2018) on every pair. On the Conneau et al. test set, it scored 0.342 vs. 0.313 MRR for EN-ET and 0.289 vs. 0.261 for EN-HR.
  • POS tagging setup: taggers were evaluated on 17 languages, with projection built on Efmaral alignments and the Plank et al. (2016) tagger.
  • POS tagging results: JW300 PROJ reached 80.8% mean accuracy, compared with 73.4 for BIBLE and 81.4 for DsDs. Using 42 source languages from the JW300–UD v2.3 overlap raised it to 81.6.

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: Natural language processing

Topics: Part-of-speech tagging

Cited by

174 papers of our library cite it (of the 140673 whose bibliographies we read); 11 of them have a page here, 163 not yet.

Cite

The paper itself, not this page. BibTeX from the library's record of the work. DOI 10.18653/v1/p19-1310

@inproceedings{agic2019jw300,
  title = {{JW300: A Wide-Coverage Parallel Corpus for Low-Resource Languages}},
  author = {Agić, Željko and Vulić, Ivan},
  booktitle = {Proceedings of the 57th Annual Meeting of the Association for Computational Linguistics},
  year = {2019},
  doi = {10.18653/v1/p19-1310},
  url = {https://doi.org/10.18653/v1/p19-1310}
}

References

39 references, 9 of them with a card of ours. ↗ leads to the work, «find» searches for one we have no id for yet.

  1. Željko Agić, Dirk Hovy, and Anders Søgaard. 2015. If all you have is a bit of the Bible: Learning POS taggers for truly low-resource languages. In Proceedings of ACL, pages 268–272. find
  2. Željko Agić, Anders Johannsen, Barbara Plank, Héctor Martı́nez Alonso, Natalie Schluter, and Anders Søgaard. 2016. Multilingual projection for parsing truly low-resource languages. Transactions of the Association for Computational Linguistics, 4:301–312. find
  3. Željko Agić, Barbara Plank, and Anders Søgaard. 2017. Cross-lingual tagger evaluation without test data. In Proceedings of EACL, pages 248–253. find
  4. Rami Al-Rfou. 2015. Polyglot: A Massive Multilingual Natural Language Processing Pipeline. Ph.D. thesis, Stony Brook University. find
  5. David Alvarez-Melis and Tommi Jaakkola. 2018. Gromov-Wasserstein alignment of word embedding spaces. In Proceedings of EMNLP, pages 1881– 1890. find
  6. Mikel Artetxe, Gorka Labaka, and Eneko Agirre. 2018. A robust self-learning method for fully unsupervised cross-lingual mappings of word embeddings. In Proceedings of ACL, pages 789–798. find our card
  7. Piotr Bojanowski, Edouard Grave, Armand Joulin, and Tomas Mikolov. 2017. Enriching word vectors with subword information. Transactions of the Association for Computational Linguistics, 5:135–146. find
  8. Tolga Bolukbasi, Kai-Wei Chang, James Zou, Venkatesh Saligrama, and Adam Kalai. 2016. Man is to computer programmer as woman is to homemaker? Debiasing word embeddings. In Proceedings of NIPS, pages 4356–4364. find our card
  9. Aylin Caliskan, Joanna J. Bryson, and Arvind Narayanan. 2017. Semantics derived automatically from language corpora contain human-like biases. Science, 356(6334):183–186. find
  10. Xilun Chen and Claire Cardie. 2018. Unsupervised multilingual word embeddings. In Proceedings of EMNLP, pages 261–270. find
  11. Christos Christodouloupoulos and Mark Steedman. 2015. A massively parallel corpus: The bible in 100 languages. Language Resources and Evaluation, 49(2):375–395. find our card
  12. Alexis Conneau, Guillaume Lample, Marc’Aurelio Ranzato, Ludovic Denoyer, and Hervé Jégou. 2018. Word translation without parallel data. In Proceedings of ICLR. find
  13. Dipanjan Das and Slav Petrov. 2011. Unsupervised part-of-speech tagging with bilingual graph-based projections. In Proceedings of ACL, pages 600–609. find
  14. Rebecca Hwa, Philip Resnik, Amy Weinberg, Clara Cabezas, and Okan Kolak. 2005. Bootstrapping parsers via syntactic projection across parallel texts. Natural Language Engineering, 11(3):311–325. find
  15. Armand Joulin, Piotr Bojanowski, Tomas Mikolov, Hervé Jégou, and Edouard Grave. 2018. Loss in translation: Learning bilingual word mapping with a retrieval criterion. In Proceedings of EMNLP, pages 2979–2984. find our card
  16. Christo Kirov, Ryan Cotterell, John Sylak-Glassman, Géraldine Walther, Ekaterina Vylomova, Patrick Xia, Manaal Faruqui, Sebastian Mielke, Arya D McCarthy, Sandra Kübler, et al. 2018. Unimorph 2.0: Universal morphology. arXiv preprint arXiv:1810.11101. find
  17. Philipp Koehn. 2005. Europarl: A parallel corpus for statistical machine translation. In Proceedings of MT Summit, volume 5, pages 79–86. find
  18. Fethi Lamraoui and Philippe Langlais. 2013. Yet another fast, robust and open source sentence aligner: Time to reconsider sentence alignment? In Proceedings of the XIV Machine Translation Summit. find
  19. Anne Lauscher and Goran Glavaš. 2019. Are we consistently biased? multidimensional analysis of biases in distributional word vectors. In Proceedings of *SEM, pages 85–91. find
  20. Pierre Lison and Jörg Tiedemann. 2016. OpenSubtitles2016: extracting large parallel corpora from movie and tv subtitles. In Proceedings of LREC. find our card
  21. Thomas Mayer and Michael Cysouw. 2014. Creating a massively parallel bible corpus. In Proceedings of LREC. find
  22. Joakim Nivre et al. 2018. Universal Dependencies 2.3. LINDAT/CLARIN digital library at the Institute of Formal and Applied Linguistics (ÚFAL), Faculty of Mathematics and Physics, Charles University. find
  23. Sunipa Dev and Jeff Phillips. 2019. Attenuating bias in word vectors. In Proceedings of AISTATS. find
  24. Robert Östling and Jörg Tiedemann. 2016. Efficient word alignment with Markov Chain Monte Carlo. Prague Bulletin of Mathematical Linguistics, 106:125–146. find
  25. Goran Glavaš, Robert Litschko, Sebastian Ruder, and Ivan Vulić. 2019. How to (properly) evaluate cross-lingual word embeddings: On strong baselines, comparative analyses, and some misconceptions. CoRR, abs/1902.00508. find
  26. Sebastian Padó and Mirella Lapata. 2009. Cross-lingual annotation projection for semantic roles. Journal of Artificial Intelligence Research, 36:307– 340. find
  27. Hila Gonen and Yoav Goldberg. 2019. Lipstick on a pig: Debiasing methods cover up systematic gender biases in word embeddings but do not remove them. In Proceedings of NAACL-HLT, pages 609–614. find our card
  28. Yedid Hoshen and Lior Wolf. 2018. Non-adversarial unsupervised word translation. In Proceedings of EMNLP, pages 469–478. find
  29. Slav Petrov, Dipanjan Das, and Ryan McDonald. 2011. A universal part-of-speech tagset. arXiv preprint arXiv:1104.2086. find
  30. Barbara Plank and Željko Agić. 2018. Distant supervision from disparate sources for low-resource part-of-speech tagging. In Proceedings of EMNLP, pages 614–620. find
  31. Barbara Plank, Sigrid Klerke, and Željko Agić. 2018. The best of both worlds: Lexical resources to improve low-resource part-of-speech tagging. arXiv preprint arXiv:1811.08757. — [arXiv:1811.08757](https://arxiv.org/abs/1811.08757) find
  32. Barbara Plank, Anders Søgaard, and Yoav Goldberg. 2016. Multilingual part-of-speech tagging with bidirectional long short-term memory models and auxiliary loss. In Proceedings of ACL, pages 412–418. find our card
  33. Sebastian Ruder, Ivan Vulić, and Anders Søgaard. 2017. A survey of cross-lingual word embedding models. arXiv preprint arXiv:1706.04902. find
  34. Samuel L. Smith, David H.P. Turban, Steven Hamblin, and Nils Y. Hammerla. 2017. Offline bilingual word vectors, orthogonal transformations and the inverted softmax. In Proceedings of ICLR. find our card
  35. Anders Søgaard, Sebastian Ruder, and Ivan Vulić. 2018. On the limitations of unsupervised bilingual dictionary induction. In Proceedings of ACL, pages 778–788. find our card
  36. Oscar Täckström, Dipanjan Das, Slav Petrov, Ryan Mc-Donald, and Joakim Nivre. 2013. Token and type constraints for cross-lingual part-of-speech tagging. Transactions of the Association for Computational Linguistics, 1:1–12. find
  37. Jörg Tiedemann. 2012. Parallel data, tools and interfaces in OPUS. In Proceedings of LREC, pages 2214–2218. find
  38. Jörg Tiedemann. 2014. Rediscovering annotation projection for cross-lingual parser induction. In Proceedings of COLING, pages 1854–1864. find
  39. David Yarowsky, Grace Ngai, and Richard Wicentowski. 2001. Inducing multilingual text analysis tools via robust projection across aligned corpora. In Proceedings of NAACL-HLT, pages 1–8. find