TY - GEN
T1 - OntoInsight - A Metric-Guided Tool for Ontology Quality Evaluation with LLM-Powered Recommendations
AU - Sammi, Daksh
AU - Bhushan, Lakshay
AU - Mutharaju, Raghava
AU - Shimizu, Cogan
N1 - © The Author(s), under exclusive license to Springer Nature Switzerland AG 2026.
PY - 2025
Y1 - 2025
N2 - Ontologies are foundational to conceptual modeling and semantic systems across diverse domains, yet evaluating and improving their quality remains a complex challenge. Existing tools often focus on syntactic correctness or complex metric reporting, lacking actionable and interpretable feedback, which is not very intuitive. We present an ontology quality evaluation tool, named OntoInsight, that caters to different types of users, from beginners to advanced, with custom recommendations, basic (simple suggestions), and advanced (involving deep technical insights) recommendations. It can handle ontologies of varying size with full ontology evaluation and modular evaluation (useful for large and complex ontologies). The pipeline automates all the stages in the tool, from metric computation (via frameworks such as OQuaRE) and seed-term-based modularization to controlled natural language (CNL) translation and targeted prompt generation for Large Language Models (LLMs). The user has the freedom to configure their own LLM API key and choose the type of evaluation and suggestions they want, according to their needs and expertise. The source code of OntoInsight is available under Apache 2.0 license at https://github.com/kracr/onto-insight.
AB - Ontologies are foundational to conceptual modeling and semantic systems across diverse domains, yet evaluating and improving their quality remains a complex challenge. Existing tools often focus on syntactic correctness or complex metric reporting, lacking actionable and interpretable feedback, which is not very intuitive. We present an ontology quality evaluation tool, named OntoInsight, that caters to different types of users, from beginners to advanced, with custom recommendations, basic (simple suggestions), and advanced (involving deep technical insights) recommendations. It can handle ontologies of varying size with full ontology evaluation and modular evaluation (useful for large and complex ontologies). The pipeline automates all the stages in the tool, from metric computation (via frameworks such as OQuaRE) and seed-term-based modularization to controlled natural language (CNL) translation and targeted prompt generation for Large Language Models (LLMs). The user has the freedom to configure their own LLM API key and choose the type of evaluation and suggestions they want, according to their needs and expertise. The source code of OntoInsight is available under Apache 2.0 license at https://github.com/kracr/onto-insight.
KW - AI-Assisted Modeling
KW - Computational Tool
KW - Controlled Natural Language
KW - Empirical Evaluation
KW - Large Language Models
KW - Modularization
KW - Ontology Quality
KW - Quality Paradigms and Metrics
UR - https://corescholar.libraries.wright.edu/cse/688
UR - https://www.scopus.com/pages/publications/105020665310
UR - https://www.scopus.com/pages/publications/105020665310#tab=citedBy
U2 - 10.1007/978-3-032-08623-5_21
DO - 10.1007/978-3-032-08623-5_21
M3 - Conference contribution
AN - SCOPUS:105020665310
SN - 978-3-032-08622-8
T3 - Lecture Notes in Computer Science
SP - 393
EP - 411
BT - Conceptual Modeling, ER 2025
A2 - Bork, Dominik
A2 - Lukyanenko, Roman
A2 - Sadiq, Shazia
A2 - Bellatreche, Ladjel
A2 - Pastor, Oscar
PB - Springer Science and Business Media Deutschland GmbH
T2 - 44th International Conference on Conceptual Modeling, ER 2025
Y2 - 20 October 2025 through 23 October 2025
ER -