Exploring Large Language Models for Ontology Alignment

Yuan He, Jiaoyan Chen, Hang Dong, Ian Horrocks
9/12/2023
cs.AIcs.CLcs.LG

Official code: KRR-Oxford/LLMap-Prelim (5 stars, Apache-2.0, last commit 2023-09-13).

Abstract

This work investigates the applicability of recent generative Large Language Models (LLMs), such as the GPT series and Flan-T5, to ontology alignment for identifying concept equivalence mappings across ontologies. To test the zero-shot performance of Flan-T5-XXL and GPT-3.5-turbo, we leverage challenging subsets from two equivalence matching datasets of the OAEI Bio-ML track, taking into account concept labels and structural contexts. Preliminary findings suggest that LLMs have the potential to outperform existing ontology alignment systems like BERTMap, given careful framework and prompt design.

View on arXivView PDF

Code Implementations(1)

KRR-Oxford/LLMap-PrelimOfficial100%
Apache-2.0

A preliminary investigation for ontology alignment (OM) with large language models (LLMs).

52Python, Jupyter NotebookJul 3, 20233 years agoApache-2.0
chatgptflan-t5large-language-modelsontologiesontology-alignment

Similar Papers

GenOM: Ontology Matching with Description Generation and Large Language Model

86% match2025

Improving LLM-based Ontology Matching with fine-tuning on synthetic data

84% match2025

OAEI-LLM-T: A TBox Benchmark Dataset for Understanding Large Language Model Hallucinations in Ontology Matching

83% match2025

Do LLMs Dream of Ontologies?

82% match1 repo12024

Complex Ontology Matching with Large Language Model Embeddings

82% match2025

Concept Space Alignment in Multilingual LLMs

82% match2024

Cite this paper

@article{he2023exploring,
  title  = {Exploring Large Language Models for Ontology Alignment},
  author = {Yuan He and Jiaoyan Chen and Hang Dong and Ian Horrocks},
  year   = {2023},
  eprint = {2309.07172},
  archivePrefix = {arXiv},
  url    = {http://arxiv.org/abs/2309.07172v1}
}

Add a badge to your README

Shows whether official code is linked for this paper.

Code of Paper badge
[![Code of Paper](https://codeofpaper.com/badge/2309.07172.svg)](https://codeofpaper.com/paper/2309.07172)

Discussion