We introduce a model of fuzzy logic programming in a truth functional fuzzy logic with arbitrary and/or tunable t-operators. This t-operator tuning is the subject of different learning from neural networks to evolutionary calculation. The choice of an operator mostly depends on the real world problem modeled, often depending on user environments and/or stereotypes. To model aggregations of different witnesses, our rules have body in disjunctive normal form. We develop fuzzy fixpoint theory and show soundness and completeness of our semantics. To control calculational efficiency, we introduce a cut with threshold. For knowledge mining and tuning of the t-operator, we restrict the problem to finding a tnorm fitting finitely many values. We show that our model of fuzzy logic programs semantically coincides with a fuzzy controller model.
f-DLPs: Extending Description Logic Programs with Fuzzy Sets and Fuzzy Logic
