Neural Description Logic Reasoning over Incomplete Knowledge Bases

Louis Mozart KAMDEM TEYOU, Luke Friedrichs, N'Dah Jean Kouagou, Caglar Demir, Yasir Mahmood, Stefan Heindorf, Axel-Cyrille Ngonga Ngomo

Keywords: concept learning, description logic, knowledge bases, neural reasoner, embeddings, SROIQ, atomic concepts

TL;DR: This paper introduces a neural reasoner that leverages embeddings to approximate symbolic reasoning in description logic, enabling concept learning from incomplete, and erroneous knowledge bases.

Abstract: Concept learning exploits background knowledge in the form of description logic axioms to learn explainable classification models from knowledge bases. Despite recent breakthroughs in the runtime of concept learners, most approaches still cannot be deployed on real-world knowledge bases. This is due to their use of description logic reasoners, which do not scale to large datasets. Moreover, these reasoners are not robust against inconsistencies and erroneous data, both being hallmarks of real datasets. We address this challenge by presenting a novel neural reasoner dubbed \approach. Our reasoner relies on embeddings to rapidly approximate the results of a symbolic reasoner. We show that our reasoner solely requires retrieving instances for atomic concepts and existential restrictions to retrieve the instances of any concept in $\mathcal{SROIQ}$. Importantly, our experiments also suggest that our reasoner is robust against missing and erroneous data.

Primary Area: other topics in machine learning (i.e., none of the above)

Code Of Ethics: I acknowledge that I and all co-authors of this work have read and commit to adhering to the ICLR Code of Ethics.

Submission Guidelines: I certify that this submission complies with the submission instructions as described on https://iclr.cc/Conferences/2025/AuthorGuide.

Anonymous Url: I certify that there is no URL (e.g., github page) that could be used to find authors’ identity.

No Acknowledgement Section: I certify that there is no acknowledgement section in this submission for double blind review.

Submission Number: 9958