This dataset is comprised of BanglaMultiHate dataset. The BanglaMultiHate dataset consists of public comments on videos collected from YouTube covering various topics, including politics, sports, and international news. This dataset is manually annotated in multi-tasks setup, including hate type, hate severity, and hate to whom.
This combination offers a rich, diverse, and detailed landscape for studying hatespeech analysis across various contexts for Bangla language.
The BanglaMultiHate are licensed under a Creative Commons Attribution-NonCommercial-ShareAlike 4.0 International License.
You should have received a copy of the license along with this work. If not, see http://creativecommons.org/licenses/by-nc-sa/4.0/.