Fumbling in Babel: An Investigation into ChatGPT's Language Identification Ability
Fuente:
arXiv
Guardado en:
| Autores principales: | , , , , |
|---|---|
| Formato: | Preprint |
| Publicado: |
2023
|
| Materias: | |
| Acceso en línea: | |
| Etiquetas: |
Agregar Etiqueta
Sin Etiquetas, Sea el primero en etiquetar este registro!
|
| _version_ | 1866913305958809600 |
|---|---|
| author | Chen, Wei-Rui Adebara, Ife Doan, Khai Duy Liao, Qisheng Abdul-Mageed, Muhammad |
| author_facet | Chen, Wei-Rui Adebara, Ife Doan, Khai Duy Liao, Qisheng Abdul-Mageed, Muhammad |
| contents | ChatGPT has recently emerged as a powerful NLP tool that can carry out a variety of tasks. However, the range of languages ChatGPT can handle remains largely a mystery. To uncover which languages ChatGPT `knows', we investigate its language identification (LID) abilities. For this purpose, we compile Babel-670, a benchmark comprising 670 languages representing 24 language families spoken in five continents. Languages in Babel-670 run the gamut from the very high-resource to the very low-resource. We then study ChatGPT's (both GPT-3.5 and GPT-4) ability to (i) identify language names and language codes (ii) under zero- and few-shot conditions (iii) with and without provision of a label set. When compared to smaller finetuned LID tools, we find that ChatGPT lags behind. For example, it has poor performance on African languages. We conclude that current large language models would benefit from further development before they can sufficiently serve diverse communities. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2311_09696 |
| institution | arXiv |
| publishDate | 2023 |
| record_format | arxiv |
| spellingShingle | Fumbling in Babel: An Investigation into ChatGPT's Language Identification Ability Chen, Wei-Rui Adebara, Ife Doan, Khai Duy Liao, Qisheng Abdul-Mageed, Muhammad Computation and Language ChatGPT has recently emerged as a powerful NLP tool that can carry out a variety of tasks. However, the range of languages ChatGPT can handle remains largely a mystery. To uncover which languages ChatGPT `knows', we investigate its language identification (LID) abilities. For this purpose, we compile Babel-670, a benchmark comprising 670 languages representing 24 language families spoken in five continents. Languages in Babel-670 run the gamut from the very high-resource to the very low-resource. We then study ChatGPT's (both GPT-3.5 and GPT-4) ability to (i) identify language names and language codes (ii) under zero- and few-shot conditions (iii) with and without provision of a label set. When compared to smaller finetuned LID tools, we find that ChatGPT lags behind. For example, it has poor performance on African languages. We conclude that current large language models would benefit from further development before they can sufficiently serve diverse communities. |
| title | Fumbling in Babel: An Investigation into ChatGPT's Language Identification Ability |
| topic | Computation and Language |
| url | https://arxiv.org/abs/2311.09696 |