MEDAI对话语料库(MEDIC)
收藏arXiv2024-01-12 更新2024-08-06 收录
下载链接：
http://arxiv.org/abs/2310.12489v3
下载链接
链接失效反馈官方服务：
资源简介：
MEDAI对话语料库(MEDIC)是由国立理工学院计算研究中心创建的一个数据集，旨在支持零样本分类任务，特别是在医疗咨询中区分医生和AI的响应。该数据集包含5640条记录，分为三个子集，分别对应医生的原始响应、ChatGPT生成的文本以及医生响应的改写版本。数据集的创建过程涉及收集和处理医疗咨询中的对话数据，旨在帮助开发更准确的文本分类方法，以区分医生和AI系统在医疗咨询中的响应。该数据集的应用领域主要集中在医疗文本分类，旨在解决如何准确识别和分类医疗咨询中医生和AI生成的文本的问题，从而提高医疗服务的透明度和信任度。
MEDAI Dialogue Corpus (MEDIC) is a dataset developed by the Computational Research Center of the National Polytechnic Institute, designed to support zero-shot classification tasks, particularly for differentiating between physician and AI-generated responses in medical consultations. It consists of 5,640 records split into three subsets, which respectively correspond to original physician responses, texts generated by ChatGPT, and rewritten versions of physician responses. The dataset was created by collecting and processing conversational data from medical consultations, with the goal of advancing the development of more accurate text classification methods to distinguish between physician and AI-generated responses in medical consultation contexts. Its primary application domain is medical text classification, aiming to address the challenge of accurately identifying and classifying texts produced by physicians and AI systems in medical consultations, thereby enhancing the transparency and trustworthiness of medical services.
提供机构：
国立理工学院计算研究中心
创建时间：
2023-10-19
