基于大语言模型的智能预问诊应用-预问诊语料数据集

1. 通过对人卫版（第9版）医学教材、诊疗指南、专家共识等专业文献的深度挖掘，细致梳理了其中记载的各类常见疾病。
2. 运用数据清洗技术剔除冗余与错误信息，最终提炼出疾病名称以及就诊科室间的精准关联。
3. 针对各类常见疾病及常见问题，根据预问诊信息采集侧重点，使用大模型生成标准的医生提问内容。

1. Conducted in-depth mining on professional literature including 9th edition medical textbooks, diagnostic and therapeutic guidelines, and expert consensuses published by People's Medical Publishing House (PMPH), and systematically sorted out all types of common diseases recorded in these materials.
2. Applied data cleaning techniques to remove redundant and erroneous information, and finally extracted precise associations between disease names and their corresponding outpatient consultation departments.
3. Generated standard physician inquiry content using Large Language Models (LLMs) for various common diseases and clinical problems, based on the key focuses of pre-consultation information collection.

搜集汇总

数据集介绍

特点

该数据集是一个针对智能预问诊应用的语料数据集，包含14000条经过清洗和标准化的医学问题，适用于各级医疗机构，用于辅助医生进行预问诊和治疗。