Common External Ontologies

An essential aspect of a build methodology is to re-use “standard” ontologies as much as possible, or what we call "core" ontologies. (See further the Lightweight, Domain Ontologies Development Methodology document.) These core ontologies of course differ by the domain and problem at hand. Some of the criteria that make for being a "core" ontology include universality, quality, highly used, broad committee or community support, well done and documented, and easily understood.

Standard "core" ontologies that we recommend are:
 * Dublin Core - basic, common metadata
 * DC Terms - a useful expansion of Dublin Core
 * schema.org - a collection of simple vocabularies for marking up Web pages recognized by leading search engines; see also its OWL ontology
 * GoodRelations - a basic e-commerce vocabulary related to products and services and the business entities that provide them and the terms of the offerings
 * Event - generic events vocabulary
 * DBpedia - extensive set of concepts and entities drawn from Wikipedia
 * FOAF - common person and contact vocabulary
 * GeoNames - standard vocabulary for places and locations
 * SKOS - simplified vocabulary for organizational and knowledge schema
 * VoID - basic datasets and linkage vocabulary, and
 * UMBEL - a set of about 26,000 'core' real-world reference concepts; another 2,000 concepts are in a 'geo' module.

Though less universal, there are also a number of secondary ontologies you may want to also consider for your own ontology development efforts:


 * BIBO - a bibliographic and citation vocabulary
 * DOAP - a project description vocabulary
 * Timeline - a temporal timeline backbone, often used with Events, and
 * SIOC - a vocabulary for describing Web- and social-related resources.

These are then supplemented with quality domain-specific ontologies, if such exist. Only then should you assign new name spaces for any newly generated ontology(ies).