LLM-supported collaborative ontology design for data and knowledge management platforms.

Area of Science:

• Materials Science
• Data Science
• Computational Science

Background:

• Managing large, diverse scientific data is challenging, hindering interoperability and progress.
• Existing ontology engineering frameworks require significant expert time and effort.
• The HyWay project focuses on hydrogen-materials interactions, requiring robust data management.

Purpose of the Study:

• To present an extended ontology engineering framework integrating Large Language Models (LLMs) for accelerated development.
• To develop a domain-specific ontology for hydrogen-materials interactions.
• To integrate the ontology into a Data and Knowledge Management Platform (DKMP) for FAIR data compliance.

Main Methods:

• Extension of the NeOn iterative ontology engineering framework.
• Integration of Large Language Models (LLMs) for task acceleration.
• Application within the HyWay project to develop the Hydrogen-Material Interaction Ontology (HMIO).
• Integration of HMIO into a Data and Knowledge Management Platform (DKMP).

Main Results:

• Development of the Hydrogen-Material Interaction Ontology (HMIO) covering 29 experimental and 14 simulation types.
• Successful integration of HMIO into the DKMP, enabling automated generation of FAIR data entry forms.
• Demonstration of HMIO compliance by design for captured data.

Conclusions:

• A hybrid human-machine workflow for ontology engineering is effective and efficient.
• This approach provides a scalable solution for creating and operationalizing complex scientific ontologies.
• The methodology advances data-driven research in materials science and other complex domains.