faisal4590aziz/bangla-health-related-paraphrased-dataset

BanglaHealthParaphrase 是一个专门为健康领域 curated 的孟加拉语同义句数据集，包含超过20万对句子，每对句子由一个原始孟加拉语句子和它的同义句组成。该数据集通过一个多步骤的管道创建，涉及从孟加拉语新闻来源中提取健康相关内容、基于英语 pivot 的同义句生成和回译，以确保语言多样性同时保持语义完整性。这个数据集旨在支持孟加拉语同义句生成、文本简化、聊天机器人开发和特定领域的NLP应用研究。

BanglaHealthParaphrase is a Bengali paraphrasing dataset specifically curated for the health domain, containing over 200,000 sentence pairs, each consisting of an original Bengali sentence and its paraphrased version. The dataset was created through a multi-step pipeline involving extraction of health-related content from Bengali news sources, English pivot-based paraphrasing, and back-translation to ensure linguistic diversity while preserving semantic integrity. This dataset aims to support research in Bengali paraphrase generation, text simplification, chatbot development, and domain-specific NLP applications.