5.3 Demonstrating OWL Inference

	There are three levels of inference to consider: no inference, RDFS, and the OWL inference provided by Pellet.
	The applications use various inference levels in action on the knowledge model we just went through.
	The application takes four parameters: an input file, the input file format (N3, RDF/XML, N-Triple, or Turtle), an output file, and the inference level (none, rdfs, or owl). When it runs,

		It loads the input file into a Jena model and applies the specified inference level to the knowledge model.
		Then it prints a summary of each individual and those statements in the model that describe it.
		The application takes an input ontology and outputs the individuals in the knowledge model, including any statements about them, after one of the three inference levels has been applied.
		The first thing the application does is extract the parameters from the arguments to the main() method.
		Next, it creates an input stream for the input file and a print writer for the output file. Once all the setup is complete, the interesting part begins.
		Depending on the inference mode, the application creates the Jena OntModel in one of three ways.
		Finally, after creating the model in one of the three ways, we load the ontology into the model using model.read(...).
		Inference occurs automatically, so no explicit action has to be taken to ensure that entailments are derived.
		The final part of the application involves iterating over the list of all individuals in the model, and printing them and all statements for which they are the subject in a summary form to the output file.

5.3 Demonstrating OWL Inference

Now we’re ready to run the application in each of the three modes, observing and placing side by side the consequences of each inference level on the information in the knowledge model.

		Performing No Inference.
		Performing RDFS Inference.
		Performing OWL Inference.

	RDFS introduces the taxonomic structure of properties and classes, allowing you to take advantage of automatic propagation of statements and class membership.
	OWL adds a lot more expressivity using restrictions and advanced class and property descriptions.