Efficient Rule Induction by Ignoring Pointless Rules

The goal of inductive logic programming (ILP) is to find a set of logical rules that generalises training examples and background knowledge. We introduce an ILP approach that identifies pointless rules. A rule is pointless if it contains a redundant literal or cannot discriminate against negative examples. We show that ignoring pointless rules allows an ILP system to soundly prune the hypothesis space. Our experiments on multiple domains, including visual reasoning and game playing, show that our approach can reduce learning times by 99% whilst maintaining predictive accuracies. Code - https://github.com/logicand-learning-lab/aaai26- implications * These authors contributed equally. This rule is too general because it entails two negative examples (f(3) and f( 9 )). Therefore, a learner can ignore its generalisations. A learner will not necessarily ignore its specialisations because one could still be helpful, such as: This rule entails all the positive and none of the negative examples. However, r 2 and r 3 are pointless because odd(A) implies int(A), so we can remove int(A) whilst preserving semantics. We call such rules reducible rules. Now consider the rule: