AgentMap framework unifies ontology matching with LLM-based multi-agent system

Researchers have developed AgentMap, a novel framework for ontology matching that unifies the discovery of both equivalence and subsumption relationships. This multi-agent system leverages Large Language Models (LLMs) to integrate semantic retrieval, hierarchical search, and collaborative reasoning. AgentMap aims to identify either the exact equivalent concept or the most specific subsumer within a target ontology for a given source concept. The framework has demonstrated promising performance in hybrid settings and outperforms specialized equivalence or subsumption matching baselines in their respective tasks. AI

IMPACT This research could improve the accuracy and scope of knowledge graph construction and semantic interoperability.

RANK_REASON This is a research paper detailing a new framework and task for ontology matching. [lever_c_demoted from research: ic=1 ai=1.0]

Read on arXiv cs.AI →

• AgentMap
• Large Language Model

• paper
• other

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