Building Rules on top of Ontologies? Inductive Logic Programming can help!

Acquiring and maintaining Semantic Web rules is very demanding and can be automated though partially by applying Machine Learning algorithms. In this paper we show that the form of Machine Learning known under the name of Inductive Logic Programming (ILP) can help. In particular, we take a critical look at two ILP proposals based on knowledge representation frameworks that integrate Description Logics and Horn Clausal Logic and draw from them general conclusions that can be considered as guidelines for further ILP research of interest to the Semantic Web.