Relational knowledge is a very common form of knowledge, which allows direct inferences to be drawn from a premise to a conclusion. This paper focuses on the problem of representing and using relational knowledge affected by uncertainty. We first discuss the intuitive meaning of the uncertainty that may affect relational knowledge and we distinguish between A-uncertainty, concerning the applicability of a relation, and V-uncertainty, concerning the validity of a relation. Then we show how the difference between A-uncertainty and V-uncertainty has received so far only limited attention in various literature proposals. Finally, we introduce an original approach to deal with uncertain relational knowledge based on AV-quantified production rules, and we discuss its main features.
Publisher's on-line resources