Neurosymbolic system combines multiple rules for conflict resolution
Ontology-Mediated Neurosymbolic Constraint Acquisition from Multiple Stakeholders
Artificial Intelligence
Summary
Many computer systems need to combine rules from different people and machines, but these rules can sometimes clash. The authors created a system that uses an ontology—a structured way to organize information—to gather and check these rules, including preferences described by AI assistants and hard limits from hardware. Their system finds conflicts and explains them so users can adjust the rules together with AI help. If conflicts remain, the system resolves them by deciding which rules are more important. They tested this on a small power grid project and showed it could work for other cases with many stakeholders.
What this means in practice
- •For systems engineers: Integrate diverse physical and user constraints into a unified model to detect and resolve conflicts in complex system design.
- •For energy management teams: Coordinate hardware limits and user preferences when planning microgrid operations to improve reliability and adaptability.
Authors
Stefan Bischof, Juliana Kainz, Danilo Valerio
Abstract
Neurosymbolic research typically assumes a pre-existing symbolic specification, leaving the upstream challenge of acquiring and formalizing requirements and constraints largely unaddressed. We present an architecture that fills this gap by using an OWL configuration ontology to mediate between neural constraint sources and downstream consumers. In this framework, LLM assistants elicit soft stakeholder preferences, while hardware specifications define hard physical and engineering limits. The ontology unifies these heterogeneous inputs, leverages description logic to identify unsatisfiability, and generates symbolic explanations that enable LLMs to interactively renegotiate terms with users. Any remaining conflicts are resolved downstream via priority-based relaxation. We illustrate our approach on a microgrid use case from the FLEXI project and argue its generalizability to multi-stakeholder domains where constraint acquisition is distributed across human and automated sources of unequal authority.