Ontologies are the newest label attached to knowledge organization systems (KOSs). They are generally specific modules and models developed by the knowledge management community. Stanford University has developed an ontology software called Protégé, which people seem to either love or hate. There doesn’t seem to be a lot of middle ground. They are either very fond of it: “Oh, I just love Protégé!”; or they hate it: “I don’t understand that tool; it doesn’t make any sense to me.” I don’t know which class you will fall into, but it is open source software and it is available. So, try it out and make your own decision.

The idea of an ontology is that it represents complex relationships among the objects or concepts themselves. So the intent of an ontology is to provide a way to connect objects.

The Physics-Astronomy Ontology Trust is looking at a way to build one KOS that can be applied to a lot of different collections —  everything from the journal literature to pre-print services to exploring the heavens and then applying those conceptual groupings to the information collection from NASA and from other places that look to the sky.

What we find is that “ontology” is a word like “facets”; it is searching for a real definition. Until the various knowledge and research communities fully coalesce, I don’t think we will know exactly what someone means by “ontology”. That means when you start a conversation with someone about an ontology it is best to first set the syntactical framework of your conversation. What do they mean when they say ontology?

Ontologies tend to include a lot of fairly complex rules governing the relationships. When people talk about semantic networks, they are not necessarily talking about ontologies. They could be speaking of applying controlled vocabularies, another kind of KOS, to their data. Ontologies may be used to describe things and might not be invoked as ontologies in a semantic web type environment. The authors might have a completely different implementation in mind entirely.

There is a definition from W3C explaining what they mean by an ontology, but there is also a definition in Wikipedia that is written completely differently. I recommend, same as with “facets”, find out what people actually mean before you get too far down a discussion road using multiple definitions for the same term – you probably need a thesaurus.

Data Harmony Thesaurus Master software follows the national and international standards for KOS creation. One of the exports for your KOS is an OWL (Web Ontology Language) Full. That is, a well-formed thesaurus that follows the standards will parse in XML as an OWL Full ontology.

Marjorie M.K. Hlava
President, Access Innovations

Note: The above posting is one of a series based on a presentation, The Theory of Knowledge, given at the Data Harmony Users Group meeting in February of 2011. The presentation covered the theory of knowledge as it relates to search and taxonomies.