What Is Metadata?

Metadata is data that describes other data (Merriam-Webster). That is the whole definition. It is the information you attach to a web page, a document, or a file so the thing inside can be found, reused, and judged relevant. It is part of any information architecture, because without it you are searching the file itself.

Think of it as a short account of what the data is. For a document: author, file size, creation date, and keywords. For a music file: artist, album, and the year of release. The same pattern shows up in information systems, social media, websites, software, music services, and online retail. You can set it by hand, choosing what to record, or generate it from the data.

Types of metadata

The uses fall roughly into business, technical, and operational. The properties differ:

• Descriptive metadata: title, subject, genre, author, creation date.
• Rights metadata: copyright status, rights holder, license terms.
• Technical metadata: file type, size, creation date and time, compression. This is what digital-object management and interoperability use.
• Preservation metadata, used when you need to navigate a collection: an item's place in a hierarchy or sequence.
• Markup languages carry metadata for navigation and interoperability: a heading, a name, a date, a list, a paragraph.

Instruments

A metadata approach starts with a metadata set: what each element is, why you use it, and which kinds of information it applies to.

Much of that is a selection list, a controlled vocabulary. The twelve months. A list of audiences. A fixed list stops typos and lets you tag quickly.

When the list grows and splits into main and subcategories, use a taxonomy: a hierarchy of nodes. Continents, countries, and cities. Main processes and sub-processes, or themes and sub-themes in an organization. A taxonomy makes a result relevant and gives you something to filter on.

Websites use facets. A facet is a group of metadata fields used together, mainly to filter.

A thesaurus organizes terms by meaning: synonyms, related terms, a definition, and a preferred term. You can also record a hierarchy in it.

If those instruments are too small, an ontology is the next step. It is a set of entities connected by whatever relationships you need. That makes it powerful and complex. Building one takes a specialist, plus the tools and standards that go with the work.

Standards

Use standards that already exist and that other organizations already apply. You avoid inventing a private scheme, and the information stays findable, reusable, and relevant because the other side uses the same terms.