An Ontology of Time

As an essential dimension of our information spacetime plays a very important role in every aspect ofour lives. A specification of temporal information isnecessarily required for a large group ofapplications including the Semantic Web and naturallanguage. In response to this need we have developeda rich ontology of temporal concepts OWL-Time(formerly DAML-Time) for describing the temporalcontent of Web pages and the temporal properties ofWeb services. Since most of the information on theWeb is in natural language it can also be used toincrease temporal awareness for natural languageapplications. The ontology is represented infirst-order logic and the OWL Web Ontology Language.It covers a very rich set of temporal conceptsextending Hobbs (2002)'s work with more complextemporal phenomena such as temporal aggregatestemporal arithmetic mixing months and days and vagueevent durations. Since missing explicit and exactdurations is one of the most common sources ofincomplete information for temporal reasoning wehave constructed an annotated corpus and appliedmachine learning techniques to automatically extractimplicit and vague event durations from text.