Does Typological Blinding Impede Cross-Lingual Sharing?

Research output: Chapter in Book/Report/Conference proceedingArticle in proceedingsResearchpeer-review

Bridging the performance gap between high- and low-resource languages has been the focus of much previous work. Typological features from databases such as the World Atlas of Language Structures (WALS) are a prime candidate for this, as such data exists even for very low-resource languages. However, previous work has only found minor benefits from using typological information. Our hypothesis is that a model trained in a cross-lingual setting will pick up on typological cues from the input data, thus overshadowing the utility of explicitly using such features. We verify this hypothesis by blinding a model to typological information, and investigate how cross-lingual sharing and performance is impacted. Our model is based on a cross-lingual architecture in which the latent weights governing the sharing between languages is learnt during training. We show that (i) preventing this model from exploiting typology severely reduces performance, while a control experiment reaffirms that (ii) encouraging sharing according to typology somewhat improves performance.
Original languageDanish
Title of host publicationProceedings of The 16th Conference of the European Chapter of the Association for Computational Linguistics (EACL 2021)
PublisherAssociation for Computational Linguistics
Publication statusAccepted/In press - 2021
EventThe 16th Conference of the European Chapter
of the Association for Computational Linguistics: EACL 2021
-
Duration: 21 Apr 202123 Apr 2021
Conference number: 16
https://2021.eacl.org/

Conference

ConferenceThe 16th Conference of the European Chapter
of the Association for Computational Linguistics
Nummer16
Periode21/04/202123/04/2021
Internetadresse

Links

ID: 260407759