Dataset Rich Results

Dataset structured data makes your data discoverable in Google Dataset Search — a dedicated search engine for researchers, journalists, and data scientists.

Table of contents

1. Required Fields
2. Recommended Fields
3. Basic Dataset
4. Research Dataset with DataCatalog
5. Common Mistakes

---

Required Fields

Field	Type	Notes
name	string	Dataset name
description	string	What the dataset contains

Recommended Fields

Field	Type	Notes
url	string	Landing page URL
sameAs	string \| string[]	DOI, repository URL
license	string	License URL
creator	Person \| Organization	Who created it
publisher	Organization	Who published it
datePublished	string	When first published
dateModified	string	Last update date
version	string	Dataset version
keywords	string[]	Descriptive keywords
distribution	DataDownload[]	Where to get the data
variableMeasured	string \| PropertyValue[]	What variables are included
spatialCoverage	string \| Place	Geographic scope
temporalCoverage	string	Time period covered
measurementTechnique	string	How data was collected
includedInDataCatalog	DataCatalog	Parent catalog

---

Basic Dataset

import { createDataset } from 'schemaorg-kit';

const dataset = createDataset({
  name: 'Global Carbon Emissions by Country 1990–2024',
  description: 'Annual CO₂ and greenhouse gas emissions data for 195 countries from 1990 to 2024. Includes total emissions, per-capita figures, and sector breakdowns.',
  url: 'https://climate-data.org/datasets/carbon-emissions',
  license: 'https://creativecommons.org/licenses/by/4.0/',
  creator: {
    '@type': 'Organization',
    name: 'Climate Research Institute',
    url: 'https://climate-data.org',
  },
  datePublished: '2025-01-15',
  dateModified: '2025-03-01',
  version: '4.2',
  keywords: ['climate', 'CO2', 'emissions', 'greenhouse gas', 'carbon footprint'],
  temporalCoverage: '1990/2024',
  spatialCoverage: {
    '@type': 'Place',
    name: 'Global',
  },
  distribution: [
    {
      '@type': 'DataDownload',
      encodingFormat: 'text/csv',
      contentUrl: 'https://climate-data.org/datasets/carbon-emissions.csv',
    },
    {
      '@type': 'DataDownload',
      encodingFormat: 'application/json',
      contentUrl: 'https://climate-data.org/datasets/carbon-emissions.json',
    },
    {
      '@type': 'DataDownload',
      encodingFormat: 'application/vnd.openxmlformats-officedocument.spreadsheetml.sheet',
      contentUrl: 'https://climate-data.org/datasets/carbon-emissions.xlsx',
    },
  ],
  variableMeasured: [
    { '@type': 'PropertyValue', name: 'total_emissions_mt', description: 'Total annual emissions in megatons CO₂e' },
    { '@type': 'PropertyValue', name: 'per_capita_tons',    description: 'Per capita emissions in metric tons CO₂' },
    { '@type': 'PropertyValue', name: 'sector',             description: 'Emission source sector (energy, transport, industry, etc.)' },
  ],
  measurementTechnique: 'Satellite observation combined with national inventory reports (UNFCCC submission data)',
  sameAs: 'https://doi.org/10.5281/zenodo.1234567',
});

---

Research Dataset with DataCatalog

import { createDataset } from 'schemaorg-kit';

const dataset = createDataset({
  name: 'ImageNet-1M — Cleaned Subset',
  description: 'A curated 1-million image subset of ImageNet with corrected labels and balanced class distribution.',
  creator: { '@type': 'Organization', name: 'ML Research Lab', url: 'https://mllab.edu' },
  datePublished: '2024-09-01',
  license: 'https://creativecommons.org/licenses/by-nc/4.0/',
  distribution: [
    {
      '@type': 'DataDownload',
      encodingFormat: 'application/zip',
      contentUrl: 'https://mllab.edu/datasets/imagenet-1m-clean.zip',
      contentSize: '45GB',
    },
  ],
  includedInDataCatalog: {
    '@type': 'DataCatalog',
    name: 'ML Research Lab Open Datasets',
    url: 'https://mllab.edu/datasets',
  },
  sameAs: 'https://paperswithcode.com/dataset/imagenet-1m-clean',
});

---

Common Mistakes

Always include at least one distribution entry with contentUrl and encodingFormat. This is what allows Dataset Search to link directly to downloadable data.

Use sameAs to link to the DOI (Digital Object Identifier) if your dataset has one — this is how academic search engines identify your dataset.

temporalCoverage uses ISO 8601 interval notation: "1990/2024" means 1990 through 2024. For an ongoing dataset: "2020/.."