Dynamic Retrieval Augmented Generation of Ontologies using Artificial Intelligence (DRAGON-AI)

A - Papers appearing in refereed journals

Toro, S., Anagnostopoulos, A. V., Bello, S. M., Blumberg, K., Cameron, R., Carmody, L., Diehl, A. D., Dooley, D. M., Duncan, W. D., Fey, P., Gaudet, P., Harris, N. L., Joachimiak, M. P., Kiani, L., Lubiana, T., Munoz-Torres, M. C., O‘Neil, S., Osumi-Sutherland, D., Puig-Barbe, A., Reese, J. T., Reiser, L., Robb, S. M., Ruemping, T., Seager, J., Sid, E., Stefancsik, R., Weber, M., Wood, V., Haendel, M. A. and Mungall, C. J. 2024. Dynamic Retrieval Augmented Generation of Ontologies using Artificial Intelligence (DRAGON-AI). Journal of Biomedical Semantics. 15 (19). https://doi.org/10.1186/s13326-024-00320-3

AuthorsToro, S., Anagnostopoulos, A. V., Bello, S. M., Blumberg, K., Cameron, R., Carmody, L., Diehl, A. D., Dooley, D. M., Duncan, W. D., Fey, P., Gaudet, P., Harris, N. L., Joachimiak, M. P., Kiani, L., Lubiana, T., Munoz-Torres, M. C., O‘Neil, S., Osumi-Sutherland, D., Puig-Barbe, A., Reese, J. T., Reiser, L., Robb, S. M., Ruemping, T., Seager, J., Sid, E., Stefancsik, R., Weber, M., Wood, V., Haendel, M. A. and Mungall, C. J.
Abstract

Background

Ontologies are fundamental components of informatics infrastructure in domains such as biomedical, environmental, and food sciences, representing consensus knowledge in an accurate and computable form. However, their construction and maintenance demand substantial resources and necessitate substantial collaboration between domain experts, curators, and ontology experts.

We present Dynamic Retrieval Augmented Generation of Ontologies using AI (DRAGON-AI), an ontology generation method employing Large Language Models (LLMs) and Retrieval Augmented Generation (RAG). DRAGON-AI can generate textual and logical ontology components, drawing from existing knowledge in multiple ontologies and unstructured text sources.

Results

We assessed performance of DRAGON-AI on de novo term construction across ten diverse ontologies, making use of extensive manual evaluation of results. Our method has high precision for relationship generation, but has slightly lower precision than from logic-based reasoning. Our method is also able to generate definitions deemed acceptable by expert evaluators, but these scored worse than human-authored definitions. Notably, evaluators with the highest level of confidence in a domain were better able to discern flaws in AI-generated definitions. We also demonstrated the ability of DRAGON-AI to incorporate natural language instructions in the form of GitHub issues.

Conclusions

These findings suggest DRAGON-AI's potential to substantially aid the manual ontology construction process. However, our results also underscore the importance of having expert curators and ontology editors drive the ontology generation process.

KeywordsOntologies; Large language models; Biocuration; Artificial intelligence; Knowledge graphs; Ontology engineering
Year of Publication2024
JournalJournal of Biomedical Semantics
Journal citation15 (19)
Digital Object Identifier (DOI)https://doi.org/10.1186/s13326-024-00320-3
PubMed ID39415214
PubMed Central ID11484368
Web address (URL)https://jbiomedsem.biomedcentral.com/articles/10.1186/s13326-024-00320-3
Open accessPublished as ‘gold’ (paid) open access
FunderNational Institutes of Health
Wellcome Trust
Publisher's version
Copyright license
CC BY 4.0
Supplemental file
Copyright license
CC BY 4.0
Output statusPublished
Publication dates
Online17 Oct 2024
Publication process dates
Accepted08 Sep 2024
PublisherBiomed Central Ltd
ISSN2041-1480

Permalink - https://repository.rothamsted.ac.uk/item/99211/dynamic-retrieval-augmented-generation-of-ontologies-using-artificial-intelligence-dragon-ai

42 total views
9 total downloads
4 views this month
0 downloads this month
Download files as zip