Great idea to use an Events ontology as the backbone to data and literature driven hypothesis evaluation. This paper discusses the use of temporal constraints, data constraints and existence constraints expressed in an ontology to validate or refute the various events that make up a hypothesis. In case of a contradiction observed from existing data in databases or literature, the system proposes alternatives to the event in question by using "neighborhood" evaluations. Very neat! References Mark's work from 2002 as a possible solution to the problem of integrating databases with differing schemas.