Recently at work I learned of an initiative called the Basic Formal Ontology. It’s an interesting take on ontology development that has captured my interest, at least for the moment.

When I learned ontology development, it took a concept-centric approach. That is, the ontology developer created classes that represented concepts of interest in the domain being modeled, created data properties that define the attributes of each class, and created object properties that describe the relationships between classes.  To take a simple example, an ontology describing Navy ships might be broken down into combatant and noncombatant classes, and this is perfectly valid provided it meets the needs of the ontology developer and the eventual system user.

In contrast, BFO takes a very different approach. BFO is focused on modeling reality–that is, those things that objectively exist in time and space. If one wanted to model Navy ships, then the model might include watercraft or ship classes, but it would not include terms like “combatant” as classes because “combatant” is a role that a ship assumes, it is not a thing in and of itself.

It’s certainly a different way of looking at ontologies. Much of this is because the developer of BFO, Dr. Barry Smith, is a philosopher by trade and training, and not an information systems specialist. He runs the National Center for Ontological Research (NCOR), which is a project of State University of New York at Buffalo. He specifically developed BFO to support scientific research, hence the focus on modeling reality instead of someone’s perception of a particular domain.

I’m not 100% certain that BFO is a better way than concept-based ontologies. But I think I see some intriguing possibilities, so I’m going to dig into it and see how things go. BFO has a lot of traction within the biomedical community, so that tells me someone thinks it has value. It’s a bit tricky to wrap one’s head around it at first, but now that I’m getting used to it I think I’m on the verge of actually being productive with it. An invaluable resource in understanding the BFO approach to ontology development is the book Building Ontologies with Basic Formal Ontology from MIT Press.

I guess I’ll see if BFO is worth the trouble or not.