Meet Dr. Amelia Kahn

Advisor: Dr. Barry Smith

Institution: State University of New York at Buffalo 

Bio:  Amelia is an IC Post-Doc working on representing probabilistic and uncertain information in ontologies.  She is based in the philosophy department at the University at Buffalo where she works with Prof. Barry Smith.  She received her Ph.D. in philosophy from the University of Texas at Austin in 2021.  Her dissertation was about applying idealized epistemological models to normal human agents.

Abstract:  Ontologies are extensively utilized within the IC for organizing data and ensuring that databases populated and maintained by different organizations can be integrated without inconsistency or error.  However, OWL, the de facto standard ontology language, struggles to adequately represent uncertain and statistical information.  This produces two serious problems for efforts within the IC to integrate and organize data using ontologies.

 First, inconsistent definitions of terms for degrees and types of likelihood are a major source of miscommunication and lack of interoperability between systems maintained by different organizations.  For instance, an event that has a 43% chance of occurring would be described as "unlikely" according to Office of the Director of National Intelligence guidelines, "likely" according to Department of Defense guidelines, and "even chance" according to NATO.  This mismatch leads to miscommunication, failures of data aggregation, and inconsistent inference drawn from the same data. This can lead to shortfalls in the quality and timeliness of decision-making.

Second, ontology-driven inferences are crucial in getting the most out of available information and enabling effective querying over databases. The inability of OWL resources to express uncertain and statistical information precludes the use of ontology-driven inferences over information of these sorts.  This drastically limits the capacity to access and utilize such information easily in intelligence analysis.