OWL as a Target for Information Extraction Systems

Current information extraction systems can do a good job of discovering entities, relations and events in natural language text. The traditional out- put of such systems is XML, with the ACE Pilot Format (APF) schema as a common target. We are developing a system that will take the output of an information extraction system as APF documents and directly populate a knowledge base with the information extracted. We report on an initial OWL ontology that covers the APF schema, a simple program to convert a set of APF documents to RDF data and a demonstration system build with Exhibit to view the results.