New method enables neuro-symbolic learning over OWL 2 DL ontologies

Researchers have developed a novel method for neuro-symbolic learning that integrates OWL 2 DL ontologies with differentiable circuits. This approach, named Baobab, compiles ontologies into Sentential Decision Diagrams (SDDs) to enable training perception networks. The system successfully uses a CNN to recognize MNIST digits, demonstrating the ability to recover latent ontology concepts and mitigate reasoning shortcuts in non-Horn description logics. AI

IMPACT This research could advance the integration of symbolic reasoning with deep learning models, potentially improving AI's ability to handle complex knowledge bases.

RANK_REASON The cluster contains a research paper detailing a novel method for neuro-symbolic learning. [lever_c_demoted from research: ic=1 ai=1.0]

Read on arXiv cs.AI →

• CNN
• EL++
• Lean 4 Programming Language
• MNIST database
• OWL 2 DL
• Sentential Decision Diagram
• SROIQ

• paper
• other

AI-generated summary · Google Gemini · from 1 sources. How we write summaries →