Unsupervised Part-Of-Speech Induction For Assamese Language Using Hierarchical Distributional Clustering
DOI:
https://doi.org/10.29284/vdyxyk25Keywords:
Unsupervised POS Tagging, Assamese Language, Distributional Clustering, Brown Clustering, Low Resource NLP.Abstract
Part-of-Speech (POS) tagging is a core task in Natural Language Processing (NLP) and underpins a wide range of downstream applications such as syntactic parsing, information extraction, and machine translation. Despite the success of supervised POS taggers for high-resource languages,
their reliance on large annotated corpora makes them impractical for low-resource languages such as Assamese.
This paper presents a fully unsupervised framework for POS induction in Assamese based on hierarchical distributional clustering inspired by Brown clustering. Words are represented using local contextual co-occurrence statistics derived from bigram and trigram contexts and are initially treated as singleton clusters. These clusters are iteratively merged using a likelihood-based objective until no further improvement is possible. Gold-standard POS annotations are used exclusively for evaluation via majority cluster mapping. Experimental results show that the proposed approach effectively induces major syntactic categories, including nouns, verbs, pronouns, and conjunctions, with high accuracy, demonstrating the practical viability of distributional clustering for linguistic analysis in low-resource settings.
References
1. P. F. Brown, P. V. deSouza, R. L. Mercer, V. J. Della Pietra, and J. C. Lai, “Class-based n-gram models of natural language,” Computational Linguistics, vol. 18, no. 4, pp. 467–479, 1992.
2. S. Goldwater and T. L. Griffiths, “A fully bayesian approach to unsupervised part-of-speech tagging,” in Proceedings of the 45th Annual Meeting of the Association for Computational Linguistics (ACL), 2007.
3. T. Berg-Kirkpatrick, A. Bouchard-Côté, J. DeNero, and D. Klein, “Painless unsupervised learning with features,” in Proceedings of the 2010 Conference of the North American Chapter of the Association for Computational Linguistics (NAACL), 2010.
4. C. Christodoulopoulos, S. Goldwater, and M. Steedman, “Two decades of unsupervised POS induction: How far have we come?” in Proceedings of the 2010 Conference on Empirical Methods in Natural Language Processing (EMNLP), 2010.
5. A. Clark, “Combining distributional and morphological information for part-of-speech induction,” in Proceedings of the 10th Conference of the European Chapter of the Association for Computational Linguistics (EACL), 2003.
6. C. D. Manning, “Part-of-speech tagging from 97% to 100%: Is it time for some linguistics?” Computational Intelligence and Knowledge Engineering / CICLing (as cited in manuscript), 2011,
7. O. Kashefi, “Unsupervised part-of-speech induction,” arXiv preprint arXiv:1801.03564, 2018.
8. D. Baishya and R. Baruah, “Highly efficient parts of speech tagging in low resource languages with improved hidden markov model and deep learning,” International Journal of Advanced Computer Science and Applications (IJACSA), vol. 12, no. 10, pp. 82–93, 2021.
9. D. Cutting, J. Kupiec, J. Pedersen, and P. Sibun, “A practical part-of-speech tagger,” in Proceedings of the Third Conference on Applied Natural Language Processing. Association for Computational Linguistics, 1992, pp. 133–140.
10. T. Brants, “Tnt: A statistical part-of-speech tagger,” Proceedings of the Sixth Applied Natural Language Processing Conference, pp. 224–231, 2000.
11. A. Ratnaparkhi, “A maximum entropy model for part-of-speech tagging,” in Proceedings of the Conference on Empirical Methods in Natural Language Processing, 1996
12. K. Toutanova, D. Klein, C. D. Manning, and Y. Singer, “Feature-rich part-of-speech tagging with a cyclic dependency network,” in Proceedings of NAACL-HLT, 2003, pp. 173–180.
13. D. Jurafsky and J. H. Martin, Speech and Language Processing. Pearson Education, 2000.
14. J. Van Gael, A. Vlachos, and Z. Ghahramani, “The infinite hmm for unsupervised pos tagging,” in Proceedings of EMNLP, 2009.
Downloads
Published
Issue
Section
License
This work is licensed under a Creative Commons Attribution 4.0 International License.
This work is licensed under a Creative Commons Attribution 4.0 International License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
