Papers by Sandesh Kumar
Corpora Generation for Urdu Grammatical Error Correction (2026.findings-acl)
Copied to clipboard
| Challenge: | grammatical error correction (GEC) for Urdu remains under-researched due to lack of annotated datasets. |
| Approach: | They propose a method for synthesizing a large dataset by collecting errors from the Urdu WikiEdits history and learning from them. |
| Outcome: | The proposed method synthesizes a large dataset and fine-tunes models against it. |