Short answer
An ontology is a model of concepts and the relationships between them, used to connect assets, products, people and places. Ontologies enable search across related concepts beyond a single hierarchy. An ontology adds value once a DAM must answer questions that span two or more entity types, such as products, campaigns and talent contracts.
How an ontology works in a DAM
An ontology works in a DAM by turning important concepts into records with typed links between them. A taxonomy says a product belongs to a product line. An ontology also says the product appears in a campaign, the campaign runs in a region, and a model in the photo holds a contract.
An ontology has three building blocks:
- Classes: the entity types, such as Asset, Product, Campaign, Person and Location.
- Relationships: named links, such as 'shows product', 'part of campaign' or 'covered by contract'.
- Instances: the real records, such as one campaign or one model.
W3C publishes the core standards. RDF describes data as subject, predicate and object statements. OWL adds formal class and property definitions, and SKOS describes concept schemes such as thesauri. Most DAMs implement a lighter version through custom entity types and relation fields.
| Aspect | Taxonomy | Ontology |
|---|---|---|
| Structure | Parent and child tree | Network of typed relationships |
| Question it answers | Which category holds this asset | Which assets connect to this product, person or place |
| Typical owner | DAM librarian | Data or information architect |
Example of an ontology
Worked example: a sportswear brand links each campaign to its products, regions and talent contracts. A talent contract ends in one region on 30 June.
The DAM follows the links from the contract to the person, from the person to every asset that shows them, and from those assets to campaigns in that region. Brand managers receive one list of assets to withdraw. A taxonomy alone cannot produce that list.
Ontology in a DAM selection
An ontology is scoped at Checkpoint 2: Define, when the buyer decides which entities the DAM must hold and which stay in PIM or MDM. DAMSelection assessment: DAM and PIM form one product content supply chain, so model products once and reference them.
Ask each vendor to show these points at Checkpoint 5: Test:
- Create a custom entity type, such as Campaign, with its own fields.
- Link one asset to a product, a campaign and a person.
- Search for all assets of products in one campaign.
- Return those relationships through the API.
The PIM and DAM Fit Map shows how product records reach each DAM.
Common ontology mistakes
The most common ontology mistake is building a full model before the basic metadata schema works. Other recurring errors:
- Copying product or customer master data into the DAM, which creates a second record that drifts from PIM or MDM.
- No owner for relationships, so links break when campaigns or contracts change.
- Treating a long keyword list as an ontology.
Related terms: Taxonomy, Semantic search.
Frequently asked
What is the difference between an ontology and a taxonomy?
A taxonomy arranges values in one parent and child hierarchy. An ontology connects different entity types through named relationships, such as asset to product, product to campaign and person to contract. Most DAMs start with taxonomies and add relationships when cross-entity questions appear.
Does a mid-market DAM need an ontology?
A mid-market DAM usually needs a few relationships before a formal ontology. DAMSelection rule of thumb: link assets to products and to rights documents first. Add campaign, person and location entities when users ask cross-entity questions every week.