Advanced Belief Reasoning in Intelligence
About the project
The ABRI project at UiO runs in collaboration with the US Army Research Labs with the purpose of advancing applications of subjective logic for intelligence analysis.
ABRI focuses on theoretical studies and development of software tools for intelligence analysis that take advantage of subjective logic's ability to handle uncertain beliefs. A typical aspect of intelligence analysis is that it must deal with uncertain evidence and incomplete knowledge. Subjective logic can explicitly handle uncertainty, which enables analysts to create more realistic reasoning models that produce more informative conclusions than is otherwise possible with traditional Bayesian approaches.
The ABRI project focuses on extending the theory of subjective logic for handling uncertainty in Bayesian reasoning, and on implementing practical tools for intelligence analysis based on subjective logic.
Subjective Logic was first described by Jøsang in 1997 as a computational framework for modelling trust. Since then it has been applied by researchers and practitioners in numerous other fields, such as natural language analysis, forensics analysis, sensor networks, multi-criteria decision making, security, semantic web, and intelligence reasoning, just to name a few. Belief based reasoning with subjective provides a promising framework for modelling and analyzing real world situations that are affected by uncertain and/or incomplete information, as well as with conflicting opinions. Subjective logic provides a rich set of operators for combining belief opinions and can be used for modelling, classifying and analyzing situations involving uncertainty such as Bayesian networks and trust networks.
The ABRI project is financed by the US Army Research Labs.
In addition to the partners UiO and the US Army research Labs, the University of Padova is also involved in ABRI. Dr. Francesco Sambo from the University of Padova was guest researcher at UiO during October 2014.