Tense Error Detection Dataset for Bangla Text

Description

We develop a high-quality Bangla corpus for tense error detection. The proposed dataset enables controlled generation of grammatically correct (and incorrect ) Bangla sentences with explicit variations in tense, subject type, and politeness level. The resulting dataset provides balanced coverage of tense categories and subject–verb agreement patterns, thereby supporting reliable benchmarking and evaluation of grammar correction systems.

Categories

Licence