Preventing, detecting & revising flaws in object property expressions

Abstract

The OWL 2 DL ontology language is very expressive and has many features for declaring complex object property expressions. Standard reasoning services for OWL ontologies take these expressions as correct and according to the ontologist's intention. However, the more one can do, the higher the chance modelling flaws are introduced; hence, an unexpected or undesired classification or inconsistency in the class hierarchy may actually be due to a mistake in the `object property box', not the class axioms. We analyse the principles of subsumption in object property hierarchies, and use it to identify the types of flaws that can occur in object property expressions. We propose the compatibility services SubProS and ProChainS that check for meaningful property hierarchies and property chaining and propose how to revise a flaw. These insights can also be used to prevent flaws and to choose the best option, which we demonstrate with the chain pattern for upward and downward distributivity over parthood relations. SubProS and ProChainS were evaluated with several ontologies, which demonstrates that such flaws do exist, that they can be isolated effectively, and useful suggestions for revisions can be proposed.