DBpediaThe DBpedia Ontology stands as one of the largest and most widely used cross-domain ontologies on the Semantic Web. Derived from the structured information in Wikipedia's infoboxes, it provides a rich schema for describing people, places, organisations, creative works, species, and many other entity types. With 768 classes and approximately 3,000 properties, the ontology serves as the backbone of the DBpedia knowledge graph and is extensively linked to by other Linked Data sources worldwide.
Background
The DBpedia project began extracting structured data from Wikipedia in 2007, initially using automated heuristics to parse infobox content. With the DBpedia 3.2 release in 2008, the project introduced a new extraction approach based on hand-generated mappings from Wikipedia infoboxes to a purpose-built ontology. This shift addressed several weaknesses in Wikipedia's infobox system, including inconsistent naming across templates, duplicate infoboxes for the same concepts, and ambiguous datatypes for property values.
The hand-curated approach evolved into a community effort with the launch of the public DBpedia Mappings Wiki alongside the 3.5 release. This wiki allows anyone to define new mappings and extend the ontology with additional classes and properties. Since the 3.7 release, the ontology has been structured as a directed acyclic graph rather than a strict tree, enabling classes to have multiple superclasses -- a change driven in part by the need to map to Schema.org.
Purpose & Scope
The DBpedia Ontology provides a structured schema for the knowledge extracted from Wikipedia infoboxes. It covers:
- People — 1,592,912 instances, the largest category
- Places — 967,491 instances covering locations worldwide
- Organisations — 317,867 instances
- Creative Works — 552,115 instances including films, books, albums, software
- Species — 190,369 instances covering biological taxonomy
- General resources — over 4.8 million instances total
The ontology's 768 classes form a subsumption hierarchy, while its 3,000 properties define attributes and relationships between instances. The schema is deliberately "shallow," meaning the class hierarchy does not go very deep, favoring breadth of coverage over deeply nested specialization.
Access Methods
| Resource | Description |
|---|---|
| Databus (Development) | Daily snapshots from the Mappings Wiki |
| DBpedia Archivo | Overview and quality monitoring of the development version |
| SPARQL Endpoint | Query the ontology and instance data directly |
| OWL Download (T-Box) | Schema definition in OWL format |
| Linked Data Interface | Dereferenceable URIs for all classes and properties |
| Mappings Wiki | Community portal for contributing mappings |
Serializations & Technical Formats
The ontology namespace is http://dbpedia.org/ontology/. The schema (T-Box) is available for download in OWL format from DBpedia Archivo. Instance data (A-Box) is distributed as part of the DBpedia dataset releases, available from the DBpedia Databus in formats including N-Triples and Turtle.
Daily development snapshots are generated automatically whenever changes are made in the Mappings Wiki. Stable releases follow the monthly DBpedia dataset release cycle, with snapshot releases every three months.
Governance & Maintenance
The DBpedia Ontology is maintained by the DBpedia Association, a non-profit organization that oversees the DBpedia project. The ontology evolves through a combination of:
- Community contributions via the Mappings Wiki, where anyone can define or update mappings
- Automated daily builds that publish development snapshots to the Databus
- Monthly release cycles that produce stable versions
- Quality monitoring through DBpedia Archivo
The Association coordinates development, hosts infrastructure, and organizes community events. It also runs a PhD program and participates in Google Summer of Code.
Notable Implementations
The DBpedia knowledge graph is one of the most interlinked datasets on the Semantic Web. Notable uses include:
- Apple Siri, Google Knowledge Graph, Amazon Alexa — have used DBpedia as a data source
- Linked Data ecosystem — DBpedia serves as a central interlinking hub, with many datasets linking to DBpedia URIs
- DBpedia Spotlight — an entity linking tool that connects text to DBpedia resources
- Wikidata — shares substantial overlap, with DBpedia providing a complementary extraction-based approach
- Academic research — widely used in NLP, information extraction, and knowledge graph research
Related Standards
- Schema.org — the DBpedia Ontology includes mappings to Schema.org classes, and the DAG structure was partly motivated by this alignment