Neuro-Symbolic Hierarchical Rule Induction
This work addresses the need for interpretable rule induction in AI, though it appears incremental as it builds on existing neuro-symbolic and ILP methods.
The authors tackled the problem of solving Inductive Logic Programming (ILP) by proposing an efficient interpretable neuro-symbolic model that learns first-order rules from hierarchical meta-rules, achieving competitive performance across various tasks including ILP, visual genome, and reinforcement learning.
We propose an efficient interpretable neuro-symbolic model to solve Inductive Logic Programming (ILP) problems. In this model, which is built from a set of meta-rules organised in a hierarchical structure, first-order rules are invented by learning embeddings to match facts and body predicates of a meta-rule. To instantiate it, we specifically design an expressive set of generic meta-rules, and demonstrate they generate a consequent fragment of Horn clauses. During training, we inject a controlled \pw{Gumbel} noise to avoid local optima and employ interpretability-regularization term to further guide the convergence to interpretable rules. We empirically validate our model on various tasks (ILP, visual genome, reinforcement learning) against several state-of-the-art methods.