10th International Conference on Information Fusion, Quebec, Canada, 9-12 July, 2007 Probabilistic Ontology for Net-Centric Fusion Kathryn Blackmond Laskey C4I Center and SEOR Department George Mason University Fairfax, VA 22030 U.S.A. klaskey@gmu.edu Paulo C.G. da Costa C4I Center George Mason University Fairfax, VA 22030 U.S.A. pcosta@gmu.edu Edward J. Wright Information Extraction and Transport, Inc. 1911 N Ft Myer Dr, Suite 600 Arlington, VA, 22209 U.S.A. ewright@iet.com Kenneth J. Laskey MITRE Corporation 1 Colshire Drive McLean, VA 22030 U.S.A. klaskey@mitre.org 1 The author's affiliation with The MITRE Corporation is provided for identification purposes only, and is not intended to convey or imply MITRE's concurrence with, or support for, the positions, opinions or viewpoints expressed by the author. Abstract – In a net-centric world, systems will be required to fuse data from geographically dispersed, heterogeneous information sources operating asynchronously, to produce up-to-date, mission-relevant knowledge to inform commanders. Realizing this vision requires overcoming a number of technical challenges. Among these is the need for semantic interoperability among systems with different internal data models and vocabularies. Ontologies are seen as a key enabling technology for semantic interoperability. Although information fusion by nature involves reasoning under uncertainty, traditional ontology formalisms provide no principled means of reasoning under uncertainty. This paper proposes the use of probabilistic ontologies within a service-oriented architecture as a means to enable semantic interoperability in net-centric fusion systems. Keywords: fusion, probabilistic ontology, Multi-Entity Bayesian Networks, service-oriented architecture, semantic interoperability. 1 Introduction The rapid expansion of connectivity in the field is increasing the problem of information overload. In the race between the availability of data and decision-makers’ ability to transform it into knowledge, many methods for using our ever-growing computational power have been devised to make life easier for warriors. In spite of these efforts, there remains a heavy reliance on human cognitive processing for breaking the barrier between information and knowledge. This leads us to the question: What is missing for IT techniques to move beyond the information paradigm and begin to work under the knowledge paradigm? One widely shared response is that the solution lies with semantics. Technologies for making semantic information explicit and computationally accessible are key to effective exploitation of data from disparate sources. Shared formal semantics enables diverse systems with different internal representations to interoperate, allows communication and appropriate processing of important contextual information that affects how data should be interpreted, and provides an automated means of enforcing business rules such as access control for security. Semantic information is commonly represented as metadata, and may comprise explicit descriptive elements or links to externally defined information or models. However, simply adding metadata arbitrarily to military C2 systems would only bring the “Tower of Babel” problem to our IT resources. When heterogeneous systems need to interoperate in an open world, vocabularies that are adequate for a single stand-alone application will break down. This happens because systems developed in isolation from each other will employ different vocabularies originally tailored to different tasks. Inevitably, there is incomplete and partial overlap of terminology and concepts. Even when concepts are clearly defined, inputs available in an open-world system may be insufficient to determine which meaning is most appropriate. Ontologies to describe the individual domains are seen as a means for tackling the problem of semantic inconsistencies among distinct systems. The need for semantic interoperability is the focus of numerous efforts to enable Web Service providers to describe the properties and capabilities of their Web Services in unambiguous, computer-interpretable form (e.g., OWL-S [1], WSMO [2], SWSL [3], and SAWSDL [4]). However, while these efforts provide either semantic models or a means to relate existing semantic models, these do not allow the description authors to specify the degree to which they consider such models complete or applicable. Indeed, SAWSDL “allows multiple annotations to be associated with a given WSDL or XML Schema ... [but] no logical relationship between them is defined by this specification.” We argue that progress on net-centric fusion is hampered by the lack of support for uncertainty in