Improving Part-of-Speech Tagging by Meta-learning
Recently, we have observed a rapid progress in the state of Part of Speech tagging for Polish. Thanks to PolEval—a shared task organized in late 2017—many new approaches to this problem have been proposed. New deep learning paradigms have helped to narrow the gap between the accuracy of POS tagging methods for Polish and for English. Still, the number of errors made by the taggers on large corpora is very high, as even the currently best performing tagger reaches an accuracy of ca. 94.5%, which translates to millions of errors in a billion-word corpus.
To further improve the accuracy of Polish POS tagging we propose to employ a meta-learning approach on top of several existing taggers. This meta-learning approach is inspired by the fact that the taggers, while often similar in terms of accuracy, make different errors, which leads to a conclusion that some of the methods are better in specific contexts than the others. We thus train a machine learning method that captures the relationship between a particular tagger accuracy and language context and in this way create a model, which makes a selection between several taggers in each context to maximize the expected tagging accuracy.
KeywordsPart-of-speech tagging Meta learning Natural language processing
- 4.Kobyliński, Ł.: PoliTa: A multitagger for Polish, pp. 2949–2954. ELRA, Reykjavík (2014). http://www.lrec-conf.org/proceedings/lrec2014/index.html
- 5.Krasnowska, K.: Morphosyntactic disambiguation for Polish with bi-LSTM neural networks. In: Vetulani Google Scholar
- 7.Piasecki, M.: Polish tagger TaKIPI: rule based construction and optimisation. Task Q. 11(1–2), 151–167 (2007)Google Scholar
- 8.Przepiórkowski, A., Bańko, M., Górski, R., Lewandowska-Tomaszczyk, B. (eds.): Narodowy Korpus Jȩzyka Polskiego. Warszawa (2012)Google Scholar
- 9.Radziszewski, A.: A tiered CRF tagger for Polish. In: Bembenik, R., Skonieczny, Ł., Rybiński, H., Kryszkiewicz, M., Niezgódka, M. (eds.) Intelligent Tools for Building a Scientific Information Platform: Advanced Architectures and Solutions, vol. 467, pp. 215–230. Springer, Heidelberg (2013). https://doi.org/10.1007/978-3-642-35647-6_16CrossRefGoogle Scholar
- 10.Radziszewski, A., Acedański, S.: Taggers gonna tag: an argument against evaluating disambiguation capacities of morphosyntactic taggers. In: Sojka, P., Horák, A., Kopeček, I., Pala, K. (eds.) TSD 2012. LNCS (LNAI), vol. 7499, pp. 81–87. Springer, Heidelberg (2012). https://doi.org/10.1007/978-3-642-32790-2_9CrossRefGoogle Scholar
- 11.Radziszewski, A., Śniatowski, T.: A memory-based tagger for Polish. In: Proceedings of the LTC (2011)Google Scholar
- 12.Vetulani, Z. (ed.): Proceedings of the 8th Language and Technology Conference: Human Language Technologies as a Challenge for Computer Science and Linguistics. Poznań, Poland, 17–19 November 2017Google Scholar
- 13.Waszczuk, J.: Harnessing the CRF complexity with domain-specific constraints. The case of morphosyntactic tagging of a highly inflected language. In: Proceedings of the 24th International Conference on Computational Linguistics (COLING 2012), Mumbai, India, pp. 2789–2804 (2012)Google Scholar
- 14.Wróbel, K.: KRNNT: Polish recurrent neural network tagger. In: Vetulani Google Scholar