Artificial Clinical Reasoning Dataset for Pneumonia Diagnosis

该数据集是一个人工创建的数据集，旨在研究如何利用结构化的表格数据和类似全科医生咨询笔记的非结构化文本来自动化区分肺炎与上呼吸道感染的临床推理。数据集不仅包含结构化变量，还包含非结构化文本，其中文本信息只有部分在结构化变量中编码，同时还包括了关于患者背景和症状的额外上下文信息。该任务的目的是基于患者的症状和背景信息来诊断肺炎。

This is a manually curated dataset developed to investigate the automation of clinical reasoning for differentiating pneumonia from upper respiratory tract infections (URIs) by leveraging structured tabular data and unstructured text similar to general practitioner consultation notes. The dataset contains both structured variables and unstructured text, where only a subset of the textual information is encoded in the structured variables, alongside additional contextual information regarding patient backgrounds and symptoms. The objective of this task is to diagnose pneumonia based on the patient's symptoms and background information.