In Summary:
"An ontology defines a common vocabulary for researchers who need to share information in a domain. It includes machine-interpretable definitions of basic concepts in the domain, and relations among them."
From: "A Guide to Creating Your First Ontology" by Natalya F. Noy and Deborah L. McGuiness (Stanford University, Stanford, California) <View PDF file on WWW>
"An ontology is a formal explicit description of concepts or classes in a domain of discourse, properties of each class describing various features and attributes of the classes, and restrictions on property values. An ontology together with a set of individual instances of classes constitutes a knowledge base. In reality, there is a fine line between where the ontology ends and the knowledge base begins."
Also from: "A Guide to Creating Your First Ontology" by Natalya F. Noy and Deborah L. McGuiness (Stanford University, Stanford, California) <View PDF file on WWW>
More Detail Still:
"An important and fundamental prerequisite to using a Representation and Reasoning System (RRS - described in Knowledge Bases) is to decide how a task domain is to be described. This requires us to decide what kind of things the domain consists of, and how they are to be related in order to express task domain problems. A major impediment to addressing this task is that there is no comprehensive theory of how to appropriately conceive and express task domains.
Despite this fundamental problem, the need for the following "commitments" is recognised:
The world can be described in terms of entities (things) and relationships among entities. An ontology is a commitment to (or a formal declaration of) what exists in any particular task domain. This assumption that the world can be described in terms of things is the same that is made in natural language and logic. This isn't a strong assumption, as entities can be anything nameable, whether concrete or abstract. For example, people, colours, emotions, numbers and times can all be considered as entities. What is a "thing" is a property of an observer as much as it is a property of the world. Different observers, or even the same observer with different goals, may divide the world up in different ways.
For each task or domain, one needs to identify specific entities and relations that can be used to express what is true about the world under consideration. How one does so can profoundly affect one's ability to solve problems in that domain."
Abridged from the final part of Section 1.3 of "Computational Intelligence: A Logical Approach" by Poole, Mackworth and Goebel (Oxford University Press (1998)). The sub-section is titled "Ontology and Conceptualisation".