KROMA: Ontology Matching with Knowledge Retrieval and Large Language Models

2025-07-18

AI Summary

Existing ontology matching approaches rely either on manually crafted rules or domain-specific models with limited generalization capability, rendering them inadequate for high-accuracy, low-overhead large-scale semantic interoperability. To address this, we propose a Large Language Model (LLM)-enhanced matching framework grounded in dual bisimilarity—operating at both the conceptual and structural levels. Our method integrates Retrieval-Augmented Generation (RAG) with knowledge graph retrieval to dynamically inject structured, lexical, and definitional knowledge; introduces concept pruning and lightweight ontology refinement to reduce computational and communication overhead; and employs prompt engineering for targeted knowledge retrieval and efficient LLM invocation. Evaluated across multiple benchmark datasets, our approach consistently outperforms conventional and state-of-the-art LLM-based methods, achieving breakthroughs in matching accuracy, inference efficiency, and scalability.

Technology Category

• Machine Learning: Large Multimodal Models (LMMs)
• Knowledge Representation and Reasoning: Ontologies
• Data Mining & Knowledge Management: Linked Open Data, Knowledge Graphs & KB Completion

Application Category

• Semantics and Knowledge: Methods to enhance, augment, integrate or synergize semantic models such as knowledge graphs and LLMs
• User Modeling, Personalization and Recommendation: Large Language Models (LLM) for user modeling and recommendation
• Graph Algorithms and Modeling for the Web: Foundation models and LLMs for Web-related graphs

Abstract

Ontology Matching (OM) is a cornerstone task of semantic interoperability, yet existing systems often rely on handcrafted rules or specialized models with limited adaptability. We present KROMA, a novel OM framework that harnesses Large Language Models (LLMs) within a Retrieval-Augmented Generation (RAG) pipeline to dynamically enrich the semantic context of OM tasks with structural, lexical, and definitional knowledge. To optimize both performance and efficiency, KROMA integrates a bisimilarity-based concept matching and a lightweight ontology refinement step, which prune candidate concepts and substantially reduce the communication overhead from invoking LLMs. Through experiments on multiple benchmark datasets, we show that integrating knowledge retrieval with context-augmented LLMs significantly enhances ontology matching, outperforming both classic OM systems and cutting-edge LLM-based approaches while keeping communication overhead comparable. Our study highlights the feasibility and benefit of the proposed optimization techniques (targeted knowledge retrieval, prompt enrichment, and ontology refinement) for ontology matching at scale.

Problem

• Enhancing ontology matching with dynamic semantic context enrichment
• Optimizing performance and efficiency in ontology matching
• Outperforming classic and LLM-based ontology matching systems

Innovation

• Uses LLMs in RAG pipeline
• Integrates bisimilarity-based concept matching
• Employs lightweight ontology refinement