README.md

metadata

dataset_info:
  features:
    - name: id
      dtype: string
    - name: title
      dtype: string
    - name: funder
      dtype: string
    - name: beneficiary
      dtype: string
    - name: source_id
      dtype: string
    - name: abstract
      dtype: string
    - name: funding_scheme
      dtype: string
    - name: label
      dtype:
        class_label:
          names:
            '0': business_rnd_innovation
            '1': fellowships_scholarships
            '2': institutional_funding
            '3': networking_collaborative
            '4': other_research_funding
            '5': out_of_scope
            '6': project_grants_public
            '7': research_infrastructure
  splits:
    - name: train
      num_bytes: 3045058
      num_examples: 2386
  download_size: 1650227
  dataset_size: 3045058
configs:
  - config_name: default
    data_files:
      - split: train
        path: data/train-*

Grant Classification Dataset

This dataset contains research grant documents classified according to a custom categorization of science, technology, and innovation (STI) policy instruments.

Dataset Description

Overview

The dataset consists of research grants from various funding sources. Each grant is classified into one of 8 categories according to a taxonomy based on the OECD's categorization of STI policy instruments.

Data Sources

• Open Sources: Publicly available grant data from various sources including NIH, Kohesio, CORDIS, and others

Features

• id: Unique identifier for the grant
• title: Title of the grant
• abstract: Abstract or description of the grant
• funder: Organization providing the funding
• funding_scheme: Type of funding scheme
• beneficiary: Organization or individual receiving the funding
• source: Origin of the data (Dimensions or Open source)
• label: Classification category (target variable)

Labels

The dataset uses the following classification categories:

1. business_rnd_innovation: Direct allocation of funding to private firms for R&D and innovation activities with commercial applications
2. fellowships_scholarships: Financial support for individual researchers or higher education students
3. institutional_funding: Core funding for higher education institutions and public research institutes
4. networking_collaborative: Tools to bring together various actors within the innovation system
5. other_research_funding: Alternative funding mechanisms for R&D or higher education
6. out_of_scope: Grants unrelated to research, development, or innovation
7. project_grants_public: Direct funding for specific research projects in public institutions
8. research_infrastructure: Funding for research facilities, equipment, and resources

Statistics

• Total examples: 2386
• Class distribution:
  • business_rnd_innovation: 170 (7.1% of examples)
  • fellowships_scholarships: 342 (14.3% of examples)
  • institutional_funding: 48 (2.0% of examples)
  • networking_collaborative: 200 (8.4% of examples)
  • other_research_funding: 34 (1.4% of examples)
  • out_of_scope: 298 (12.5% of examples)
  • project_grants_public: 1157 (48.5% of examples)
  • research_infrastructure: 137 (5.7% of examples)

Usage

from datasets import load_dataset

# Load the dataset
dataset = load_dataset("SIRIS-Lab/grant-classification-dataset")

# Access the data
train_data = dataset["train"]
validation_data = dataset["validation"]
test_data = dataset["test"]

# Example of accessing a sample
sample = train_data[0]
print(f"Title: {sample['title']}")
print(f"Label: {sample['label']}")