As far as I understand it, an ontology infers constraints based on the relation-types between entities. For example, say we have the follow two
statements:
Joe must read the paper.
Joe must dispose of the paper.
Joe and paper are entities, read and dispose of are relational types.
If we dictate that Joe must perform both actions, we can infer a sequence of actions by understanding the relationships between Joe and paper. If we know from our ontology that disposal of reading material negates the ability to read the material afterwards, we know that in order for Joe to meet the first commitment (must read), it has to be performed prior to his second commitment (must dispose of).
Procedural code handles this as well, by explicitly stating the order of precendence of an operation. The above two statements, though, are just asserting two relational conditions between entities, not a sequence of operation. However, by having an ontological understanding of the relation "read" and "dispose of", we can infer the order of operation.
|