SenWiCh: Sense-Annotation of Low-Resource Languages for WiC using Hybrid Methods
Fuente:
arXiv
Saved in:
| Main Authors: | , , , , , , , , , , , , |
|---|---|
| Format: | Preprint |
| Published: |
2025
|
| Subjects: | |
| Online Access: | |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866909699618635776 |
|---|---|
| author | Goworek, Roksana Karlcut, Harpal Shezad, Muhammad Darshana, Nijaguna Mane, Abhishek Bondada, Syam Sikka, Raghav Mammadov, Ulvi Allahverdiyev, Rauf Purighella, Sriram Gupta, Paridhi Ndegwa, Muhinyia Dubossarsky, Haim |
| author_facet | Goworek, Roksana Karlcut, Harpal Shezad, Muhammad Darshana, Nijaguna Mane, Abhishek Bondada, Syam Sikka, Raghav Mammadov, Ulvi Allahverdiyev, Rauf Purighella, Sriram Gupta, Paridhi Ndegwa, Muhinyia Dubossarsky, Haim |
| contents | This paper addresses the critical need for high-quality evaluation datasets in low-resource languages to advance cross-lingual transfer. While cross-lingual transfer offers a key strategy for leveraging multilingual pretraining to expand language technologies to understudied and typologically diverse languages, its effectiveness is dependent on quality and suitable benchmarks. We release new sense-annotated datasets of sentences containing polysemous words, spanning ten low-resource languages across diverse language families and scripts. To facilitate dataset creation, the paper presents a demonstrably beneficial semi-automatic annotation method. The utility of the datasets is demonstrated through Word-in-Context (WiC) formatted experiments that evaluate transfer on these low-resource languages. Results highlight the importance of targeted dataset creation and evaluation for effective polysemy disambiguation in low-resource settings and transfer studies. The released datasets and code aim to support further research into fair, robust, and truly multilingual NLP. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2505_23714 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | SenWiCh: Sense-Annotation of Low-Resource Languages for WiC using Hybrid Methods Goworek, Roksana Karlcut, Harpal Shezad, Muhammad Darshana, Nijaguna Mane, Abhishek Bondada, Syam Sikka, Raghav Mammadov, Ulvi Allahverdiyev, Rauf Purighella, Sriram Gupta, Paridhi Ndegwa, Muhinyia Dubossarsky, Haim Computation and Language Artificial Intelligence This paper addresses the critical need for high-quality evaluation datasets in low-resource languages to advance cross-lingual transfer. While cross-lingual transfer offers a key strategy for leveraging multilingual pretraining to expand language technologies to understudied and typologically diverse languages, its effectiveness is dependent on quality and suitable benchmarks. We release new sense-annotated datasets of sentences containing polysemous words, spanning ten low-resource languages across diverse language families and scripts. To facilitate dataset creation, the paper presents a demonstrably beneficial semi-automatic annotation method. The utility of the datasets is demonstrated through Word-in-Context (WiC) formatted experiments that evaluate transfer on these low-resource languages. Results highlight the importance of targeted dataset creation and evaluation for effective polysemy disambiguation in low-resource settings and transfer studies. The released datasets and code aim to support further research into fair, robust, and truly multilingual NLP. |
| title | SenWiCh: Sense-Annotation of Low-Resource Languages for WiC using Hybrid Methods |
| topic | Computation and Language Artificial Intelligence |
| url | https://arxiv.org/abs/2505.23714 |