This paper presents a simple framework for Horn clause abduction, with probabilities associated with hypotheses. It is shown how this representation can represent any probabilistic knowledge representable in a Bayesian belief network. The main contributions are in finding a relationship between logical and probabilistic notions of evidential reasoning. This can be used as a basis for a new way to implement Bayesian Networks that allows for approximations to the value of the posterior probabilities, and also points to a way that Bayesian networks can be extended beyond a propositional language.
As the technology for building knowledge based systems has matured, important lessons have been learned about the relationship between the architecture of a system and the nature of the problems it is intended to solve. We are implementing a knowledge engineering tool called BART that is designed with these lessons in mind. BART is a Bayesian reasoning tool that makes belief networks and other probabilistic techniques available to knowledge engineers building classificatory problem solvers. BART has already been used to develop a decision aid for classifying ship images, and it is currently being used to manage uncertainty in systems concerned with analyzing intelligence reports. This paper discusses how state-of-the-art probabilistic methods fit naturally into a knowledge based approach to classificatory problem solving, and describes the current capabilities of BART.
CADRE is a system for the detection of complex events in relational data. It implements a form of abductive reasoning that combines data-driven and pattern-driven inferencing to efficiently search for matches in massive amounts of data. It has been applied to a number of pattern detection problems, most notably to the problem of threat detection in massive amounts of data. This paper describes the details of CADRE processing and compares CADRE with other systems for abductive inference. We show that CADRE has unique features that make it especially suitable for the problem of pattern detection in very large relational databases.
Decision support and information fusion in complex domains requires reasoning about inherently uncertain properties of and relationships among varied and often unknown number of entities interacting in differing and often unspecified ways. Tractable probabilistic reasoning about such complex situations requires combining efficient inference with logical reasoning about which variables to include in a model and what the appropriate probability distributions are. This paper describes the PLASMA architecture for predicate logic based assembly of situationspecific probabilistic models. PLASMA maintains a declarative representation of a decision theoretically coherent first-order probabilistic domain theory. As evidence about a situation is absorbed and queries are processed, PLASMA uses logical inference to reason about which known and/or hypothetical entities to represent explicitly in the situation model, which known and/or uncertain relationships to represent, what functional forms and parameters to specify for the local distributions, and which exact or approximate inference and/or optimization techniques to apply. We report on a prototype implementation of the PLASMA architecture within IET's Quiddity*Suite, a knowledge-based probabilistic reasoning toolkit. Examples from our application experience are discussed.