What is an Ontology?

An ontology is a representation of knowledge, generally of a particular domain, written with a standardized, structured syntax. An ontology contains concepts, also called resources, that serve to characterize the domain. An ontology can relate resources to other resources, either internally or in other ontologiesA type of relational controlled vocabulary, which provides for categories, relationships, rules and axioms among metadata elements. Typically a hierarchy of classes and terms, an ontology is a machine-readable way of relating metadata terminology. Related Guide. Resources can represent the existence of an individual entity ("grannySmithApple445" or "cellPhoneOwnedByJerry"), define subclasses that have a relationship to a class ("grannySmithApple" subclass of "apple", "cellPhone" subclass of "telephone"), and define a class to be associated with a group of entities or subclasses ("fruit," "communication device"). Resources can be associated and defined using relationshipsConnections between metadata terms within a vocabulary. These relationships can connect terms by scope, provenance, or other well-defined criteria.. For example, an individual resource is associated with a class ("apple" is a member of "fruit") or a class is associated with an ontology (class "fruit" is described in an ontology called "food").

Ontologies vs. Controlled Vocabularies

A formal specificationAny description of how to store metadata. Specifications have no limitations on the level of required documentation and no requirement for formal approval, publishing or governance by a broad community-based organization. Related Guide of a vocabulary can be something we are all familiar with: a plain list of words, a dictionaryIn the context of metadata, a dictionary is a type of controlled flat vocabulary, which provides a list of metadata terms, definitions and additional information within a specific domain. Related Guide, a taxonomy, or a thesaurusA type of relational controlled vocabulary which provides a list of terms, with specific relationships between the terms. Related Guide. Or, it can be a more technical document: an Entity-Relational (ER) diagram, an Object Model in Unified Modeling Language (UMLUnified Modeling Language ) diagram, or an eXtensible Markup Language (XMLeXtensible Markup Language ) schemaIn the context of metadata, a description of the data represented within a database.. Many other representations are possible for controlled vocabularies.

There are two views on what makes a controlled vocabulary qualify as an ontology. In the first view, simply expressing the vocabulary in an OWL file makes it an ontology, and further subtleties of classification are not important.

In the second view, a controlled vocabulary becomes an ontology when its concepts are defined explicitly and at least some of them are defined as classes. In addition to this requirement, an ontology needs to conform to strict hierarchical subclass relationships between the classes [Gruber, 1993]. The trivial ontologies that simply specify some terms do not serve any other purpose than naming, and so to many ontologists do not further the semantic webThe transformation of the web from an inherently human-interpretable medium to an inherently computer-interpretable medium. In the semantic web, machines can read and understand the content published in the network.; their lack of classes, relations, and properties make them not sufficiently powerful to be designated as ontologies.

Ontologies can include all of the following, but are not required to include them, depending on which perspective from above you adhere to:

• Classes (general things, types of things)
• Instances (individual things)
• Relationships among things
• Properties of the things
• Functions, processes, constraints, and rulesIn the context of crosswalking, rules are a process which define how to deal consistently with complex element mappings. Rules are created and applied during the mapping of elements from the source schema to the target schema, when one-to-one relationships between schema elements do not exist. Related Guide relating to the things

Other Considerations

It is worth noting that a science, also called Ontology (denoted here with a capital "O"), has existed for several centuries and has helped to inform current practice in the computer and information sciences, though major differences do exist. Ontology is the study and description of reality, or what can be said to exist, and an attempt to categorize existing things and their relationship to one another. While Ontology seeks to describe every possible thing, ontologists in computer science tend to work in particular knowledge domains, focusing their work on smaller portions of a larger ontological whole.

Some of these terms, particularly 'ontology', have been defined many different ways in different publications. Deborah McGuinness, for example, has proposed that an ontology could be construed as including the entire spectrum of controlled vocabularies. In this guide we use one of the more common definitions, but usage in other papers or contexts may vary. In "The Semantic Web", the authors refer to the 'ontology spectrum', ranging from weak semantic entities like taxonomiesA multi-level controlled vocabulary in which metadata terms are grouped according to subject-specific classes, usually hierarchical. Related Guide, to strong semantic solutions like conceptual models and advanced logics.

References:

• T. Gruber, A Translation Approach to Portable Ontology Specification., Knowledge Acquisition 5(2), 199-220, 1993.
• M. C. Daconta, L. J. Obrst, K. T. Smith, The Semantic Web. Wiley Publishing Inc. copyright 2003. (pp 156-158 ff)