Bengali Hate Speech Dataset

本研究介绍了名为‘Bengali Hate Speech Dataset’的数据集，由Shahjalal University of Science and Technology创建。该数据集包含30,000条来自YouTube和Facebook的Bengali语评论，其中10,000条标记为仇恨言论。数据集涵盖7个类别，通过众包方式标记并由专家验证。此数据集旨在支持Bengali语社交媒体中的仇恨言论检测研究，解决现有资源不足的问题。This study introduces the dataset named 'Bengali Hate Speech Dataset', which was created by Shahjalal University of Science and Technology. The dataset contains 30,000 Bengali comments sourced from YouTube and Facebook, among which 10,000 are annotated as hate speech. It covers 7 categories, with annotations completed via crowdsourcing and validated by experts. This dataset is designed to support research on hate speech detection in Bengali social media, addressing the shortage of existing relevant resources.

搜集汇总

背景与挑战

背景概述

该数据集包含30,000条孟加拉语社交媒体评论，其中10,000条被标记为仇恨言论，分为7个类别，由众包方式标记并经专家验证，旨在促进孟加拉语仇恨言论检测研究。