Integrating Healthcare Ontologies: Inconsistency Tolerance and Case Study

A major challenge for ontology integration is to effectively deal with inconsistencies that arise during the merging process. Because of the explosive nature of classical logic, the common strategy in existing merging tools is to choose between the contradictory pieces of information and maintain consistency. In many cases inconsistent information may be useful for intelligent reasoning activities. For example, in healthcare systems inconsistent information may be required to provide a full clinical perspective so any information loss is undesirable. In this paper we present a multi-valued logic based merging system that has inconsistency tolerant behavior and avoids information loss. As an application of the system in the healthcare domain, a result of merging a subset of two healthcare ontologies SNOMED CT and ICNP is presented.