TopiOCQA
收藏arXiv2025-09-30 收录
下载链接：
https://mcgill-nlp.github.io/topiocqa/
下载链接
链接失效反馈官方服务：
资源简介：
该数据集旨在支持开放领域的信息检索对话，其中代理根据维基百科的知识来回应用户的提问。每个对话轮次都关联着一个来自维基百科的、长度为200个标记的黄金段落。这一任务属于开放领域对话范畴。
This dataset is developed to support open-domain information retrieval conversational tasks, where AI agents respond to user queries based on knowledge from Wikipedia. Each dialogue turn is associated with a 200-token golden paragraph sourced from Wikipedia. This task falls within the category of open-domain dialogue tasks.
