Skip to content
Technology transformation expertise firm · ARDIANE Group Innovate, secure, succeed

Glossary

Data catalogue: definition, content and how it is populated

Domain

Data

In brief

A data catalogue is the inventory of an organisation's datasets, described by their metadata: content, structure, owner, classification and access conditions. A term of common usage, the data catalogue serves to find data, understand it and know who is accountable for it, without hosting the data itself.

Page contents

Definition of the data catalogue

A data catalogue (data catalog) is the inventory of an organisation’s datasets, described by their metadata: what each dataset contains, how it is structured, who is accountable for it, how it is classified and under what conditions it can be obtained. The catalogue points to the data without hosting it. This definition describes a usage, whose scope varies from one source to another.

The W3C gives a formulation of it for catalogues published on the Web. Its DCAT vocabulary (Data Catalog Vocabulary), whose version 3 is a recommendation of 22 August 2024, defines the catalogue as “a curated collection of metadata about resources”. It aims at interoperability between catalogues and makes it possible to describe datasets and data services.

The questions the catalogue answers

Before using data, an analyst, a project manager or a data protection officer asks the same questions. Does this data exist, and where? What does this column mean? Who is accountable for it, and who authorises access? Is it personal data?

Without a catalogue, these answers may exist only in the memory of a few people. The catalogue brings them together in a single search point, open to business users as well as technical teams.

What a catalogue entry contains, and who fills it in

Four families of metadata are distinguished here:

  • technical: schema, types, location, format, date of last update;
  • business: description of the dataset, field definitions, associated glossary terms;
  • governance: owner, steward, classification, retention period, access conditions;
  • usage and quality: views, derived datasets, results of checks when they are published.

Two ways of populating the catalogue complement each other. Connectors read the schemas of databases, warehouses and reporting tools, which keeps technical metadata up to date without manual entry. Meaning, classification and the owner’s name are declared: they fall to the data stewards, under the authority of the owners. The origin of a dataset received from outside is also declared, as explained in our article data provenance vs data lineage.

Data catalogue, glossary, lineage, record of processing activities: not to be confused

The business glossary sets the meaning of terms and indicators. The catalogue describes the datasets and links them to these terms.

Data lineage describes the links between datasets. The catalogue describes each one on its own, and can display the graph.

The reference repository resulting from master data management contains reference data. The catalogue contains metadata.

The technical catalogue of a query engine or a lakehouse registers tables so that engines know how to read them. It can feed the data catalogue, without having its documentary purpose.

The record of processing activities, provided for in Article 30 of the GDPR, Regulation (EU) 2016/679, lists processing operations on personal data and not datasets. The catalogue does not replace it.

Example: three “contracts” tables in the catalogue

Take a telecommunications operator where an analyst has to calculate a churn rate. She searches for “contract” in the catalogue, which shows her three datasets. The first is the daily extract from the management application. The second is a copy made for a past campaign, with no owner and no recent update. The third is the table that the sales department designates as the reference: the effective date of termination is defined in it, the dataset is classified as “personal data” and access is requested from its owner.

The analyst chooses this third dataset and sends her request to its owner. The catalogue told her which table to use and whom to contact. It does not say whether the values are correct, unless the results of quality checks are published in it.

Deciding which data to catalogue, who maintains the entries and with what mandate is a matter of data governance. The automatic collection of technical metadata relies on the data platform.

FAQ

Frequently asked questions

Another question? Ask us

What is a data catalogue?

A data catalogue is the inventory of an organisation's datasets, each described by its metadata: content, structure, owner, classification, access conditions. It serves to find data, understand it and identify the person accountable for it. The data catalogue points to the data without hosting it.

What metadata does a data catalogue contain?

A data catalogue entry brings together technical metadata (schema, location, update date), business metadata (description, field definitions), governance metadata (owner, classification, access conditions) and, depending on the tooling, usage and quality information. Technical metadata can be collected automatically, whereas meaning, classification and the owner's name are declared by designated people.

What is the difference between a data catalogue and a business glossary?

The business glossary sets the meaning of the organisation's terms and indicators, independently of the tables where they are stored. The data catalogue describes the datasets that exist and links their fields to the glossary terms. The two instruments complement each other, and a single tool can combine them.

What is DCAT?

DCAT (Data Catalog Vocabulary) is a W3C RDF vocabulary designed to facilitate interoperability between data catalogues published on the Web. Its version 3 is a W3C recommendation of 22 August 2024. DCAT makes it possible to describe datasets and data services according to a model shared by several catalogues.

A project to bring to completion?

A first conversation is enough to understand your challenge, identify the right people and tell you frankly what we can bring.