Neuro-Symbolic Inductive Logic Programming with Logical Neural Networks
DOI:
https://doi.org/10.1609/aaai.v36i8.20795Keywords:
Machine Learning (ML), Knowledge Representation And Reasoning (KRR), Humans And AI (HAI), Data Mining & Knowledge Management (DMKM)Abstract
Recent work on neuro-symbolic inductive logic programming has led to promising approaches that can learn explanatory rules from noisy, real-world data. While some proposals approximate logical operators with differentiable operators from fuzzy or real-valued logic that are parameter-free thus diminishing their capacity to fit the data, other approaches are only loosely based on logic making it difficult to interpret the learned ``rules". In this paper, we propose learning rules with the recently proposed logical neural networks (LNN). Compared to others, LNNs offer a strong connection to classical Boolean logic thus allowing for precise interpretation of learned rules while harboring parameters that can be trained with gradient-based optimization to effectively fit the data. We extend LNNs to induce rules in first-order logic. Our experiments on standard benchmarking tasks confirm that LNN rules are highly interpretable and can achieve comparable or higher accuracy due to their flexible parameterization.Downloads
Published
2022-06-28
How to Cite
Sen, P., Carvalho, B. W. S. R. de, Riegel, R., & Gray, A. (2022). Neuro-Symbolic Inductive Logic Programming with Logical Neural Networks. Proceedings of the AAAI Conference on Artificial Intelligence, 36(8), 8212-8219. https://doi.org/10.1609/aaai.v36i8.20795
Issue
Section
AAAI Technical Track on Machine Learning III