“4137X˙CH03˙Akerkar” 2007/9/8 10:42 page 82 #12
82 CHAPTER 3 Semantic Web
proposed by Arp´ırez and colleagues (1998) and omez-P´erez (1998). It follows the set of
principles given next.
Clarity and objectivity: An ontology should provide precise and objective definitions in
natural language form.
Completeness: Definitions should be expressed by necessary and sufficient conditions.
Coherence: An ontology should allow for consistent inferences with definitions.
Maximal monotonic extendibility: Inclusion of new general or specific terms in the on-
tology without revision of existing definitions.
Minimal ontological commitment: Make as few axioms as possible about the modeled
world.
Ontological distinction principle: Classes with different identity criteria must be disjoint.
A good practice to support the ontology building process is to produce an ontology specification
document written in natural language with information, such as the purpose of the ontology,
its end users, use case scenarios, degree of formality used to codify the ontology, and its scope.
There are other types of ontologies, which take into consideration measures such as the
formality of the language or the level of dependence on a particular task or point of view.
Top-level ontologies depict broad concepts similar to space, time, object, event or action,
which do not depend on a particular domain.
Domain ontologies and task ontologies depict the vocabulary for a generic domain (such
as bioinformatics), a task or activity, by means of specific terms.
Application ontologies depict concepts that rely on a specific domain and task. The
concepts react to roles performed by domain entities while doing certain tasks.
3.4 Ontology Languages for the Web
The World Wide Web Consortium established two standards—Resource Description Frame-
work (RDF) and the Web Ontology Language (OWL)—which define how developers specify
these relationships. These standards define how data text and facts are linked and how to de-
fine relationships between elements. The syntax used to make semantic associations is called
an RDF triple. The triple definition lets information describe itself. Applications can interpret
RDF triples without knowing many details about the source or schema of the data. Self-
description reduces the coding and data mapping needed to exchange information between
applications.
Presently, the most important ontology languages for the Web are the following:
XML presents syntax for structured documents but enforces no semantic constraints
on the meaning of these documents.
XML Schema is a language for limiting the structure of XML documents.
RDF is a data model for objects and relations between them. It provides simple seman-
tics for the data model. These data models can be represented in XML syntax.

Get Building an Intelligent Web: Theory and Practice now with O’Reilly online learning.

O’Reilly members experience live online training, plus books, videos, and digital content from 200+ publishers.