In short. DataCatalog groups datasets through dataset and includedInDataCatalog, while each Dataset carries its own description and download metadata.

Definition: Schema.org defines DataCatalog as a collection of datasets. DataCatalog structured data describes that collection and connects it to individual datasets through catalog membership properties.

Key facts

Schema.org places DataCatalog under CreativeWork, a type for creative works.
The dataset property connects a catalog to a member described as a Dataset.
The inverse property, includedInDataCatalog, connects a dataset to its containing catalog.
As of 17 October 2026, Google's Dataset structured data documentation lists includedInDataCatalog as recommended and explains that a dataset can belong to multiple repositories.

How does a catalog group published datasets?

Schema.org's Dataset definition describes a body of structured information about a topic. The catalog groups those bodies of information through explicit membership links. A catalog's dataset value describes a member; the member's includedInDataCatalog value describes its catalog.

Schema.org's catalog membership property expects a DataCatalog value on a Dataset. This gives the relationship a direction: a dataset points to the collection containing it. The inverse relationship expresses the same membership from the collection's side.

Download information belongs in the dataset's distribution property. The Dataset definition gives this property a DataDownload value describing a downloadable form at a particular location and in a particular format. It permits repeated distributions for different variations. Catalog membership and downloadable formats therefore describe separate relationships within the dataset's metadata.

What does the markup look like?

Google's published example uses JSON-LD, a format for expressing linked data as JSON, to describe the NCDC Storm Events Database. Its catalog field is:

{
  "includedInDataCatalog": {
    "@type": "DataCatalog",
    "name": "data.gov"
  }
}

This field belongs inside the example's Dataset description, whose required properties include name and description. Its distribution objects describe downloads separately.

Where should the descriptions appear?

Google's repository guidance recommends dataset structured data on each canonical landing page, the principal page describing that dataset. It distinguishes these pages from listings that group datasets.

For descriptions repeated in listings or other copies, the same guidance recommends sameAs links to the canonical dataset page. A practical publishing pattern is to describe each dataset on its landing page and record catalog membership there.

FAQ

Can my dataset belong to several catalogs?

Yes. Google's DataCatalog guidance explicitly allows the same dataset to appear in more than one repository. Catalog membership describes the collections containing it.

How do I describe several download formats?

Use the dataset's distribution property. Schema.org's Dataset definition permits repeated values for downloadable variations, with each DataDownload describing a location and format.

Sources

Check how a page reads to answer engines. Open the GEO score analyzer