For our purposes, a short and simple definition is that an ontology is a conceptual model. An ontology can be similar to a database schema or an object model, but it can also be significantly different from these more familiar forms. A slightly longer definition is this: ontologies are subject-specific vocabularies that describe and categorize entities, both tangible or intangible. For example, some ontologies describe medical entities like diseases and treatments, while others describe financial entities like stocks and bonds. (Actually, like many computer science terms, the word " ontology" has a broad meaning grounded in metaphysical philosophy. See for example this wikipedia article).

Most people are familiar with the concept of a taxonomy, which is a particular kind of limited ontology. A well known example is the categorization of biological organisms into Kingdom, Phylum, Class, Order, Family, Genus, and Species.

Ontologies are useful because they allow experts to capture a complete and unambiguous model of a domain in question, in a format accessible from any modern computing platform. With proper planning, the extension and refinement of this model can take place in parallel with development of any custom software needed to leverage the captured knowledge in an organization's processes.

frof

Certainly, people who are defing modern information systems (i.e. programmers) need ontologies for at least two things: interchange and configuration. The techniques we teach build upon the demonstrated success of XML in these areas.

But more broadly, the modeling power offered by semantic technology is potentially relevant to many tasks faced by modern knowledge workers, including researchers, analysts, legal and health care workers, librarians, designers, web developers, entrepeneurs, and engineers.

Some examples of applications where ontology integration should be considered:

• Media resource management (e.g. library science)
• Definition of message vocabularies (for B2B, B2G, G2G applications)
• Clinical medical applications (e.g. insurance and billing, drug interaction, decision support)
  
• Scientific collaboration workflows involving complex datasets
  
• Statistical population analysis (e.g. correlation of survey responses wth election data)
  
• Matchmaking (e.g. matching jobs with resumes, matching romantic profiles, etc.)
• Financial applications (forecasts, what-if analysis, pricing models)
• Product modelling (e.g. complex financial products involving securities)
• Intelligent search applications
• Rigorous validation systems
  
• Planning and optimization
  
• Simulations, games, and educational applications
• Law enforcement applications
• Information security models (e.g. complex privelege hierarchies)

To understand more about why we recommend using ontologies in preference to traditional mechanisms, please continue on to read about How RDF/OWL Ontologies Capture Durable Value for Organizations.