README.md

metadata

license: apache-2.0
configs:
  - config_name: default
    data_files:
      - split: train
        path: data/train-*
      - split: validation
        path: data/validation-*
      - split: test
        path: data/test-*
dataset_info:
  features:
    - name: pmid
      dtype: int64
    - name: journal
      dtype: string
    - name: title
      dtype: string
    - name: abstract
      dtype: string
    - name: keywords
      dtype: string
    - name: pub_type
      dtype: string
    - name: authors
      dtype: string
    - name: doi
      dtype: string
    - name: label
      sequence: int64
    - name: text
      dtype: string
  splits:
    - name: train
      num_bytes: 85014595
      num_examples: 24960
    - name: validation
      num_bytes: 9075648
      num_examples: 2500
    - name: test
      num_bytes: 21408810
      num_examples: 6239
  download_size: 63244210
  dataset_size: 115499053
task_categories:
  - text-classification
language:
  - en
size_categories:
  - 10K<n<100K

Dataset Card for Dataset Name

Dataset Description

• Homepage: BioCreative VII LitCovid Track
• Paper: Multi-label classification for biomedical literature: an overview of the BioCreative VII LitCovid Track for COVID-19 literature topic annotations

Dataset Summary

Topic annotation in LitCovid is a multi-label document classification task that assigns one or more labels to each article. There are 7 topic labels used in LitCovid: Treatment, Diagnosis, Prevention, Mechanism, Transmission, Epidemic Forecasting, and Case Report. These topics have been demonstrated to be effective for information retrieval and have also been used in many downstream applications related to COVID-19.

Dataset Structure

Data Instances and Data Splits

• the training set contains 24,960 articles from LitCovid;
• the validation set contains 6,239 articles from LitCovid;
• the test set contains 2,500 articles from LitCovid;

Data Fields

with the following fields retrieved from PubMed/LitCovid: • pmid: PubMed Identifier

• journal: journal name

• title: article title

• abstract: article abstract

• keywords: author-provided keywords

• pub_type: article type, e.g., journal article

• authors: author names

• doi: Digital Object Identifier

• label: annotated topics in list format indicating absence or presence of labels in the order 'Treatment,Diagnosis,Prevention,Mechanism,Transmission,Epidemic Forecasting,Case Report'

• text: The text field is created as follows: '[Title]: ' + title + ' [Abstract]: ' + abstract + ' [Keywords]: ' + keywords