• Article / Chapter
  • Do LSTMs really work so well for PoS tagging? – A replication study

Do LSTMs really work so well for PoS tagging? – A replication study

A recent study by Plank et al. (2016) found that LSTM-based PoS taggers considerably improve over the current state-of-the-art when evaluated on the corpora of the Universal Dependencies project that use a coarse-grained tagset. We replicate this study using a fresh collection of 27 corpora of 21 languages that are annotated with fine-grained tagsets of varying size. Our replication confirms the result in general, and we additionally find that the advantage of LSTMs is even bigger for larger tagsets. However, we also find that for the very large tagsets of morphologically rich languages, hand-crafted morphological lexicons are still necessary to reach state-of-the-art performance.

Preview

Category

Conference:
Conference on Empirical Methods in Natural Language Processing, Copenhagen, Denmark, September 7-11, 2017
Date Issued:
2017
URN:
urn:nbn:de:hbz:464-20211019-115751-4
DOI:
10.18653/v1/D17-1076
Language:
English
Type of Resource:
Text
Collection:
E-Publications
Subject categories of the Deutsche Nationalbibliografie:
004 Data processing Computer sciences
Hosting institution:
Faculty of Engineering, Computer Science and Applied Cognitive Science, Computer Science, Language Technology Lab
Info on primary publication:
Horsmann, T., Zesch, T. (2017) Do LSTMs really work so well for PoS tagging? – A replication study. In: Proceedings of the 2017 Conference on Empirical Methods in Natural Language Processing, pp. 727-736. Association for Computational Linguistics. DOI: https://doi.org/10.18653/v1/D17-1076

Cite

Citation style:
Could not load citation form.

Rights

Use and reproduction:
This work may be used under a
CC BY 4.0 LogoCreative Commons Attribution 4.0 License (CC BY 4.0)
.