VlogQA

VlogQA是一个专注于越南口语的机器阅读理解数据集，由信息技术大学创建。该数据集包含10,076个基于YouTube视频转录文档的问答对，主要涉及食物和旅行主题。数据集的创建过程包括转录收集、问答对创建、数据集修改和质量保证四个主要阶段。VlogQA数据集旨在解决越南语口语阅读理解任务中的挑战，特别是在处理日常口语中的非正式语法和方言差异方面。

VlogQA is a machine reading comprehension dataset dedicated to Vietnamese spoken language, developed by the University of Information Technology. It comprises 10,076 question-answer pairs sourced from YouTube video transcripts, with the majority focusing on food and travel themes. The construction of the VlogQA dataset involves four core stages: transcript collection, question-answer pair creation, dataset refinement, and quality assurance. This dataset is designed to tackle the challenges inherent in Vietnamese spoken language machine reading comprehension tasks, especially those related to informal grammar and dialectal discrepancies in everyday spoken communication.