This work investigates the most basic units that underlie contextualized word embeddings, such as BERT - the so-called word pieces. In Morphologically-Rich Languages (MRLs) which exhibit morphological fusion and non-concatenative morphology, the different units of meaning within a word may be fused, intertwined, and cannot be separated linearly. Therefore, when using word-pieces in MRLs, we must consider that: (1) a linear segmentation into sub-word units might not capture the full morphological complexity of words; and (2) representations that leave morphological knowledge on sub-word units inaccessible might negatively affect performance. Here we empirically examine the capacity of word-pieces to capture morphology by investigating the task of multi-tagging in Hebrew, as a proxy to evaluating the underlying segmentation. Our results show that, while models trained to predict multi-tags for complete words outperform models tuned to predict the distinct tags of WPs, we can improve the WPs tag prediction by purposefully constraining the word-pieces to reflect their internal functions. We conjecture that this is due to the naïve linear tokenization of words into word-pieces, and suggest that linguistically-informed word-pieces schemes, that make morphological knowledge explicit, might boost performance for MRLs.
|Title of host publication||SIGMORPHON 2020 - 17th SIGMORPHON Workshop on Computational Research in Phonetics Phonology, and Morphology, Proceedings of the Workshop|
|Publisher||Association for Computational Linguistics (ACL)|
|Number of pages||6|
|State||Published - 2020|
|Event||17th SIGMORPHON Workshop on Computational Research in Phonetics Phonology, and Morphology, SIGMORPHON 2020 as part of the 58th Annual Meeting of the Association for Computational Linguistics, ACL 2020 - Virtual, Online, United States|
Duration: 10 Jul 2020 → …
|Name||Proceedings of the Annual Meeting of the Association for Computational Linguistics|
|Conference||17th SIGMORPHON Workshop on Computational Research in Phonetics Phonology, and Morphology, SIGMORPHON 2020 as part of the 58th Annual Meeting of the Association for Computational Linguistics, ACL 2020|
|Period||10/07/20 → …|
Bibliographical noteFunding Information:
We thank Yoav Goldberg, Noah Smith, Omer Levy and three reviewers for interesting discussions of an earlier draft. This research is funded by an ERC Grant #677352 and an ISF grant #1739/26, for which we are grateful.
© 2020 Association for Computational Linguistics.