Probabilistic ontologies for knowledge fusion

To cope with asymmetric threats in an increasingly network centric environment, todaypsilas command support systems must interoperate with a diverse collection of other systems. As a natural consequence, the focus is changing from data fusion to knowledge fusion. This new reality creates the need for advanced techniques that exploit not only the syntactic structure of knowledge bases, but also the semantic content. Ontologies play a major role in semantically aware systems, providing a means for highly effective knowledge sharing. However, they lack a standardized treatment of uncertainty, a ubiquitous feature of multisource fusion problems. This paper discusses the applicability of Probabilistic Ontologies designed in Probabilistic OWL (PR-OWL) to address this unmet need. The potential of PR-OWL probabilistic ontologies is demonstrated through a case study in the counterterrorism domain. A simple PR-OWL ontology for the case study was implemented using a recently released PR-OWL knowledge representation and reasoning system.