LLMs in Action: Robust Metrics for Evaluating Automated Ontology Annotation Systems

Ontologies are critical for organizing and interpreting complex domain-specific knowledge, with applications in data integration, functional prediction, and knowledge discovery. As the manual curation of ontology annotations becomes increasingly infeasible due to the exponential growth of biomedical...

Full description

Saved in:
Bibliographic Details
Main Authors: Ali Noori, Pratik Devkota, Somya D. Mohanty, Prashanti Manda
Format: Article
Language:English
Published: MDPI AG 2025-03-01
Series:Information
Subjects:
Online Access:https://www.mdpi.com/2078-2489/16/3/225
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1849341633368162304
author Ali Noori
Pratik Devkota
Somya D. Mohanty
Prashanti Manda
author_facet Ali Noori
Pratik Devkota
Somya D. Mohanty
Prashanti Manda
author_sort Ali Noori
collection DOAJ
description Ontologies are critical for organizing and interpreting complex domain-specific knowledge, with applications in data integration, functional prediction, and knowledge discovery. As the manual curation of ontology annotations becomes increasingly infeasible due to the exponential growth of biomedical and genomic data, natural language processing (NLP)-based systems have emerged as scalable alternatives. Evaluating these systems requires robust semantic similarity metrics that account for hierarchical and partially correct relationships often present in ontology annotations. This study explores the integration of graph-based and language-based embeddings to enhance the performance of semantic similarity metrics. Combining embeddings generated via Node2Vec and large language models (LLMs) with traditional semantic similarity metrics, we demonstrate that hybrid approaches effectively capture both structural and semantic relationships within ontologies. Our results show that combined similarity metrics outperform individual metrics, achieving high accuracy in distinguishing child–parent pairs from random pairs. This work underscores the importance of robust semantic similarity metrics for evaluating and optimizing NLP-based ontology annotation systems. Future research should explore the real-time integration of these metrics and advanced neural architectures to further enhance scalability and accuracy, advancing ontology-driven analyses in biomedical research and beyond.
format Article
id doaj-art-96da1f46ac1344ba8c4fb2d16a43dc35
institution Kabale University
issn 2078-2489
language English
publishDate 2025-03-01
publisher MDPI AG
record_format Article
series Information
spelling doaj-art-96da1f46ac1344ba8c4fb2d16a43dc352025-08-20T03:43:36ZengMDPI AGInformation2078-24892025-03-0116322510.3390/info16030225LLMs in Action: Robust Metrics for Evaluating Automated Ontology Annotation SystemsAli Noori0Pratik Devkota1Somya D. Mohanty2Prashanti Manda3Informatics and Analytics, University of North Carolina, Greensboro, NC 27412, USAFractal Analytics, New York, NY 10006, USAUnited Health Group, Minnetonka, MN 55343, USADepartment of Computer Science, University of Nebraska, Omaha, NE 68182, USAOntologies are critical for organizing and interpreting complex domain-specific knowledge, with applications in data integration, functional prediction, and knowledge discovery. As the manual curation of ontology annotations becomes increasingly infeasible due to the exponential growth of biomedical and genomic data, natural language processing (NLP)-based systems have emerged as scalable alternatives. Evaluating these systems requires robust semantic similarity metrics that account for hierarchical and partially correct relationships often present in ontology annotations. This study explores the integration of graph-based and language-based embeddings to enhance the performance of semantic similarity metrics. Combining embeddings generated via Node2Vec and large language models (LLMs) with traditional semantic similarity metrics, we demonstrate that hybrid approaches effectively capture both structural and semantic relationships within ontologies. Our results show that combined similarity metrics outperform individual metrics, achieving high accuracy in distinguishing child–parent pairs from random pairs. This work underscores the importance of robust semantic similarity metrics for evaluating and optimizing NLP-based ontology annotation systems. Future research should explore the real-time integration of these metrics and advanced neural architectures to further enhance scalability and accuracy, advancing ontology-driven analyses in biomedical research and beyond.https://www.mdpi.com/2078-2489/16/3/225semantic similarityLLMsontology annotationgene ontology
spellingShingle Ali Noori
Pratik Devkota
Somya D. Mohanty
Prashanti Manda
LLMs in Action: Robust Metrics for Evaluating Automated Ontology Annotation Systems
Information
semantic similarity
LLMs
ontology annotation
gene ontology
title LLMs in Action: Robust Metrics for Evaluating Automated Ontology Annotation Systems
title_full LLMs in Action: Robust Metrics for Evaluating Automated Ontology Annotation Systems
title_fullStr LLMs in Action: Robust Metrics for Evaluating Automated Ontology Annotation Systems
title_full_unstemmed LLMs in Action: Robust Metrics for Evaluating Automated Ontology Annotation Systems
title_short LLMs in Action: Robust Metrics for Evaluating Automated Ontology Annotation Systems
title_sort llms in action robust metrics for evaluating automated ontology annotation systems
topic semantic similarity
LLMs
ontology annotation
gene ontology
url https://www.mdpi.com/2078-2489/16/3/225
work_keys_str_mv AT alinoori llmsinactionrobustmetricsforevaluatingautomatedontologyannotationsystems
AT pratikdevkota llmsinactionrobustmetricsforevaluatingautomatedontologyannotationsystems
AT somyadmohanty llmsinactionrobustmetricsforevaluatingautomatedontologyannotationsystems
AT prashantimanda llmsinactionrobustmetricsforevaluatingautomatedontologyannotationsystems