Eeny, meeny, miny, moe. How to choose data for morphological inflection.

Saliha Muradoglu, Mans Hulden


Abstract
Data scarcity is a widespread problem for numerous natural language processing (NLP) tasks within low-resource languages. Within morphology, the labour-intensive task of tagging/glossing data is a serious bottleneck for both NLP and fieldwork. Active learning (AL) aims to reduce the cost of data annotation by selecting data that is most informative for the model. In this paper, we explore four sampling strategies for the task of morphological inflection using a Transformer model: a pair of oracle experiments where data is chosen based on correct/incorrect predictions by the model, model confidence, entropy, and random selection. We investigate the robustness of each sampling strategy across 30 typologically diverse languages, as well as a 10-cycle iteration using Natügu as a case study. Our results show a clear benefit to selecting data based on model confidence. Unsurprisingly, the oracle experiment, which is presented as a proxy for linguist/language informer feedback, shows the most improvement. This is followed closely by low-confidence and high-entropy forms. We also show that despite the conventional wisdom of larger data sets yielding better accuracy, introducing more instances of high-confidence, low-entropy, or forms that the model can already inflect correctly, can reduce model performance.
Anthology ID:
2022.emnlp-main.492
Volume:
Proceedings of the 2022 Conference on Empirical Methods in Natural Language Processing
Month:
December
Year:
2022
Address:
Abu Dhabi, United Arab Emirates
Editors:
Yoav Goldberg, Zornitsa Kozareva, Yue Zhang
Venue:
EMNLP
SIG:
Publisher:
Association for Computational Linguistics
Note:
Pages:
7294–7303
Language:
URL:
https://aclanthology.org/2022.emnlp-main.492
DOI:
10.18653/v1/2022.emnlp-main.492
Bibkey:
Cite (ACL):
Saliha Muradoglu and Mans Hulden. 2022. Eeny, meeny, miny, moe. How to choose data for morphological inflection.. In Proceedings of the 2022 Conference on Empirical Methods in Natural Language Processing, pages 7294–7303, Abu Dhabi, United Arab Emirates. Association for Computational Linguistics.
Cite (Informal):
Eeny, meeny, miny, moe. How to choose data for morphological inflection. (Muradoglu & Hulden, EMNLP 2022)
Copy Citation:
PDF:
https://aclanthology.org/2022.emnlp-main.492.pdf