We apply methods of abduction derived from propositional probabilistic reasoning to predicate probabilistic reasoning, in particular inductive logic, by treating finite predicate knowledge bases as potentially infinite propositional knowledge bases. It is shown that for a range of predicate knowledge bases (such as those typically associated with inductive reasoning) and several key proposition...

We propose an interleaved inductive-abductive model for reasoning about complex spatio-temporal narratives. Typed Inductive Logic Programming (Typed-ILP) is used as a basis for learning the domain theory by generalising from observation data, whereas abductive reasoning is used for noisy data correction by scenario and narrative completion thereby improving the inductive learning to get semanti...

We explore the notion of duality for defeasible entailment relations induced by preference orderings on states. We then show that such preferential entailment relations may be used to characterise Peircean inductive and abductive reasoning. Interpreting the preference relations as accessibility relations establishes modular Gödel-Löb logic as the modal logic of inductive and abductive reasoning. §

Comodels of Lawvere theories, i.e. models in Set , model state spaces with algebraic access operations. Standard equational reasoning is known to be sound but incomplete for comodels. We give two sound and complete calculi for equational reasoning over comodels: an inductive calculus for equality-on-the-nose, and a coinductive/inductive calculus for equality modulo bisimulation which captures b...

My doctoral research focuses on understanding semantic knowledge in neural network models trained solely to predict natural language (referred as models, or LMs), by drawing insights from the study of concepts and categories grounded cognitive science. I propose a framework inspired 'inductive reasoning,' phenomenon that sheds light how humans utilize background make inductive leaps generalize ...

This paper focuses on problems where the reuse of old solutions seems appropriate but the conventional CBR methodology is not adequate because a complete description of the new problem is not available to trigger case retrieval. We describe an information theoretic technique that solves this problem by producing focused questions to fill out the case description. This use of information theoret...

Whether performed by national agencies or local law enforcement, the ultimate objective of intelligence analysis is to develop timely inferences that can be acted upon with confidence. To this end, effective intelligence analysis consists of integrating collected information and then developing and testing hypotheses based on that information through successive iterations of additional data col...