Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

MedDialog: Two Large-scale Medical Dialogue Datasets

Xuehai He, Shu Chen, Zeqian Ju, Xiangyu Dong, Hongchao Fang, Sicheng Wang, Yue Yang, Jiaqi Zeng, Ruisi Zhang, Ruoyu Zhang, Meng Zhou, Penghui Zhu, Pengtao Xie

TL;DR

The paper addresses the shortage of large-scale, diverse medical dialogue data needed to train effective telemedicine systems. It presents two public datasets, MedDialog-EN (English) and MedDialog-CN (Chinese), with substantial numbers of conversations and utterances across a wide range of medical specialties. EN comprises 257,454 consultations (514,908 utterances) from 2008–2020, while CN comprises 1,145,231 consultations (3,959,333 utterances; 3,209,660 after merging same-speaker turns) from 2010–2020, sourced from public platforms and organized into descriptive and dialog parts with diagnosis/treatment information where available. The work demonstrates that these are among the largest, most diverse medical dialogue resources to date, publicly available to accelerate the development of telemedicine-era dialogue systems and reduce population biases.

Abstract

Medical dialogue systems are promising in assisting in telemedicine to increase access to healthcare services, improve the quality of patient care, and reduce medical costs. To facilitate the research and development of medical dialogue systems, we build two large-scale medical dialogue datasets: MedDialog-EN and MedDialog-CN. MedDialog-EN is an English dataset containing 0.3 million conversations between patients and doctors and 0.5 million utterances. MedDialog-CN is an Chinese dataset containing 1.1 million conversations and 4 million utterances. To our best knowledge, MedDialog-(EN,CN) are the largest medical dialogue datasets to date. The dataset is available at https://github.com/UCSD-AI4H/Medical-Dialogue-System

MedDialog: Two Large-scale Medical Dialogue Datasets

TL;DR

The paper addresses the shortage of large-scale, diverse medical dialogue data needed to train effective telemedicine systems. It presents two public datasets, MedDialog-EN (English) and MedDialog-CN (Chinese), with substantial numbers of conversations and utterances across a wide range of medical specialties. EN comprises 257,454 consultations (514,908 utterances) from 2008–2020, while CN comprises 1,145,231 consultations (3,959,333 utterances; 3,209,660 after merging same-speaker turns) from 2010–2020, sourced from public platforms and organized into descriptive and dialog parts with diagnosis/treatment information where available. The work demonstrates that these are among the largest, most diverse medical dialogue resources to date, publicly available to accelerate the development of telemedicine-era dialogue systems and reduce population biases.

Abstract

Medical dialogue systems are promising in assisting in telemedicine to increase access to healthcare services, improve the quality of patient care, and reduce medical costs. To facilitate the research and development of medical dialogue systems, we build two large-scale medical dialogue datasets: MedDialog-EN and MedDialog-CN. MedDialog-EN is an English dataset containing 0.3 million conversations between patients and doctors and 0.5 million utterances. MedDialog-CN is an Chinese dataset containing 1.1 million conversations and 4 million utterances. To our best knowledge, MedDialog-(EN,CN) are the largest medical dialogue datasets to date. The dataset is available at https://github.com/UCSD-AI4H/Medical-Dialogue-System

Paper Structure

This paper contains 7 sections, 2 figures, 1 table.

Table of Contents

  1. Introduction
  2. Datasets
  3. MedDialog-EN
  4. MedDialog-CN
  5. Advantages of our datasets
  6. Related Works
  7. Conclusions

Figures (2)

  • Figure 1: An exemplar consultation, which includes (1) description of medical conditions of the patient, (2) dialogue between doctor and patient.
  • Figure 2: An exemplar consultation, which includes (1) description of medical conditions and history of the patient, (2) dialogue between doctor and patient, and (3) diagnosis and treatment suggestions given by the doctor.