Datasets:
Tasks:
Text Retrieval
Modalities:
Text
Formats:
json
Languages:
Chinese
Size:
10K - 100K
ArXiv:
License:
metadata
license: apache-2.0
task_categories:
- text-retrieval
language:
- zh
tags:
- chinese-similar-literature-retrieval
pretty_name: CMIRB Benchmark
size_categories:
- 10M<n<100M
dataset_info:
- config_name: default
features:
- name: q_id
dtype: string
- name: p_id
dtype: string
- name: score
dtype: int64
splits:
- name: test
num_bytes: 21072
num_examples: 439
- config_name: corpus
features:
- name: id
dtype: string
- name: text
dtype: string
splits:
- name: corpus
num_bytes: 28016252
num_examples: 36758
- config_name: queries
features:
- name: id
dtype: string
- name: text
dtype: string
splits:
- name: queries
num_bytes: 326559
num_examples: 439
configs:
- config_name: default
data_files:
- split: test
path: qrels/test.jsonl
- config_name: corpus
data_files:
- split: corpus
path: corpus.jsonl
- config_name: queries
data_files:
- split: queries
path: queries.jsonl
π Overview
CMIRB: Chinese Medical Information Retrieval Benchmark
CMIRB is a specialized multi-task dataset designed specifically for medical information retrieval. It consists of data collected from various medical online websites, encompassing 5 tasks and 10 datasets, and has practical application scenarios.
Name | Description | Query #Samples | Doc #Samples |
---|---|---|---|
MedExamRetrieval | Medical multi-choice exam | 697 | 27,871 |
DuBaikeRetrieval | Medical search query from BaiDu Search | 318 | 56,441 |
DXYDiseaseRetrieval | Disease question from medical website | 1,255 | 54,021 |
MedicalRetrieval | Passage retrieval dataset collected from Alibaba | 1,000 | 100,999 |
CmedqaRetrieval | Online medical consultation text | 3,999 | 100,001 |
DXYConsultRetrieval | Online medical consultation text | 943 | 12,577 |
CovidRetrieval | COVID-19 news articles | 949 | 100,001 |
IIYiPostRetrieval | Medical post articles | 789 | 27,570 |
CSLCiteRetrieval | Medical literature citation prediction | 573 | 36,703 |
CSLRelatedRetrieval | Medical similar literatue | 439 | 36,758 |
π GitHub
Github link AutoMIR
π Paper
Paper link arXiv