This document presents some simple example OWL ontologies and discusses some of the inferences that can be made about the classes and individuals in those ontologies.
This version produced 12th July, 2005. Note that the document uses mathematical symbols — these may not display correctly in Internet Explorer.
The OWL Web Ontology Language describes a language for ontologies. This language is equipped with a formal semantics described in the OWL Web Ontology Semantics and Abstract Syntax [OWL S&AS]. Using these semantics, inferences about ontologies and individuals can be made. It is not always obvious why these inferences have occurred, however. Explanation of reasoning process is a topic of research interest — however this is still to reach a state where it is effective.
The purpose of this hands on session is to explore some of the effects of applying reasoning in OWL. There are some example ontologies provided. Within these ontologies, there are a number of inferences that can be made, both about the classes defined in the ontologies (for example discovering that class definitions are inconsistent), and about the instances of the ontology (for example discovering that a particular instance is inferred to be a member of a particular class). You can use your favourite editor/reasoner to examine the inferences that can be drawn from the given assertions.
For each of these example inferences, you should look at the underlying model and try and work out why the inference is being made. Sometimes the inference may be due to some direct assertion, sometimes it may be due to the interaction between a number of assertions.
For all of the examples, feel free to experiment and change the definitions given in order to better understand what the operators mean and how they interact.
There are no solutions given for this session, although there are explanations provided for many of the inferences. Nor are you necessarily expected to work through all the examples in the time given.