Dealing with Incompleteness and Vagueness in Inductive Logic Programming

Incompleteness and vagueness are inherent properties of knowledge in several real world domains and are particularly pervading in those domains where entities could be better described in natural language. In order to deal with incomplete and vague structured knowledge, several fuzzy extensions of Description Logics (DLs) have been proposed in the literature. In this paper, we address the issues raised by incomplete and vague knowledge in Inductive Logic Programming (ILP). We present a novel ILP method for inducing fuzzy DL axioms from crisp DL assertions and discuss the results obtained in comparison with related works.