Neural Reasoning for Robust Instance Retrieval in $mathcal{SHOIQ}$

2025-10-23

AI Summary

Existing description logic (DL)-based concept learning methods rely on symbolic reasoners, exhibiting poor robustness on real-world knowledge bases due to data inconsistency, incompleteness, or errors. Method: We propose EBR, a neuralized instance retrieval framework that models DL semantics via knowledge embedding and replaces symbolic reasoning with an end-to-end differentiable neural network. EBR operates under $mathcal{SHOIQ}$ semantics and approximates arbitrary complex concepts using only atomic concepts and existential restrictions ($exists R.C$). Contribution/Results: EBR significantly improves robustness against noisy and incomplete data without requiring preprocessing (e.g., data cleaning or consistency repair). Experiments on realistic noisy knowledge bases show that EBR achieves up to 23.6% higher instance recall stability compared to state-of-the-art DL reasoners (e.g., HermiT, FaCT++), marking the first approach to enable faithful concept approximation in $mathcal{SHOIQ}$ using minimal syntactic constructs while fully bypassing traditional symbolic inference.

Technology Category

• Knowledge Representation and Reasoning: Description Logics
• Machine Learning: Statistical Relational/Logic Learning
• Data Mining & Knowledge Management: Intelligent Query Processing

Application Category

• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking
• Semantics and Knowledge: Scalable techniques for the creation, curation, publication, maintenance, and consumption of large, Web-based, structured, reusable, knowledge graphs and ontologies
• Graph Algorithms and Modeling for the Web: Graph embeddings and representation learning for Web-related graphs

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 neuro-symbolic concept learning, most approaches still cannot be deployed on real-world knowledge bases. This is due to their use of description logic reasoners, which are not robust against inconsistencies nor erroneous data. We address this challenge by presenting a novel neural reasoner dubbed EBR. Our reasoner relies on embeddings to approximate the results of a symbolic reasoner. We show that EBR solely requires retrieving instances for atomic concepts and existential restrictions to retrieve or approximate the set of instances of any concept in the description logic $mathcal{SHOIQ}$. In our experiments, we compare EBR with state-of-the-art reasoners. Our results suggest that EBR is robust against missing and erroneous data in contrast to existing reasoners.

Problem

• Developing neural reasoning for robust instance retrieval
• Addressing inconsistencies in description logic knowledge bases
• Approximating symbolic reasoning using embedding techniques

Innovation

• Neural reasoner uses embeddings to approximate symbolic reasoning
• Retrieves instances for atomic concepts and restrictions
• Robust against missing and erroneous data inconsistencies