BYTETOOLS

Dataset Schema Generator

Build Dataset JSON-LD for Google Dataset Search with licence, DOI, creator, temporal and spatial coverage, variables measured and DataDownload files.

Dataset
Licence and provenance
Coverage
Downloads (distribution)
Download 1

Dataset JSON-LD

Google Dataset Search needs at least a name and a description, and rewards a licence, an identifier such as a DOI, and at least one DataDownload with a real file URL. A distribution row is only written out once you give it a file URL.

What is the Dataset Schema Generator?

The Dataset Schema Generator builds the schema. org Dataset markup that Google Dataset Search reads.

  • Temporal coverage built as an ISO interval, open-ended if you leave one side blank
  • Bounding box written as a GeoShape in the correct coordinate order
  • Bounding-box validation catches half-filled and inverted corners
  • Repeatable DataDownload rows with encodingFormat and contentUrl
  • Common open licences as presets plus a custom URL option
  • Copy as a script block or download the raw JSON

How to use the Dataset Schema Generator

  1. 1

    Enter the dataset name, landing page URL and a description of what the data contains.

  2. 2

    Add identifiers such as a DOI, plus keywords and a licence from the list or your own URL.

  3. 3

    Fill in the creator and the data catalog the dataset is published in.

  4. 4

    Set the temporal coverage dates and, if the data is geographic, the four bounding-box coordinates.

  5. 5

    Add a download row for each file format with its file URL, then copy or download the JSON-LD.

About the Dataset Schema Generator

The Dataset Schema Generator builds the schema.org Dataset markup that Google Dataset Search reads. Enter the dataset name, description, licence and identifier and it produces the JSON-LD, including the creator, the data catalog it lives in and whether access is free.

Coverage is handled properly: a start and end date become a temporalCoverage interval, and four bounding-box coordinates become a spatialCoverage Place with a GeoShape box in the minimum-latitude, minimum-longitude, maximum-latitude, maximum-longitude order the specification defines. You can list the variables measured and add as many DataDownload distributions as you have file formats.

The tool runs entirely in your browser — no dataset, URL or description you type is uploaded anywhere. It is built for research groups, open data portals and anyone publishing data who wants it to be discoverable rather than buried.

Frequently asked questions

What does Google Dataset Search need in the markup?

At a minimum a name and a description. Beyond that, Google strongly recommends a licence, an identifier such as a DOI, the creator, and at least one distribution with a real downloadable file URL, because those are the fields people filter and judge a dataset by.

How do I write a bounding box in schema.org?

Use a Place with a geo GeoShape whose box property lists four space-separated numbers in the order minimum latitude, minimum longitude, maximum latitude, maximum longitude. This generator builds that string for you and warns if the south-west corner is not smaller than the north-east one.

What format should temporalCoverage use?

An ISO 8601 interval with a slash between the two dates, for example 2011-01-01/2016-12-31. For data that is still being collected, leave the end date blank and the tool writes an open-ended interval ending in two dots.

What is the difference between identifier and url in Dataset schema?

url is the human-readable landing page, while identifier is a persistent identifier for the data itself, usually a DOI written as a full https://doi.org/ URL. Having both lets search engines link duplicate copies of the same dataset in different repositories.

Can I mark up a dataset that is not free to access?

Yes. Untick the free-access box and isAccessibleForFree is written as false. Being honest about access helps, because Dataset Search shows the flag and users can filter on it — a paywalled dataset marked as free just produces frustrated visitors.

Related tools