A Machine-Assisted Framework for Ontology Development and Standardization: Case Study in Digital Health Technologies.

Area of Science:

• Digital Health
• Medical Informatics
• Knowledge Management

Background:

• Digital health technologies (DHTs) are transforming healthcare delivery and research.
• The rapid growth of DHT literature presents significant challenges for knowledge organization and retrieval.
• Existing ontology development methods in digital health are often manual, hindering efficiency and scalability.

Purpose of the Study:

• To propose and illustrate a novel framework for managing the expanding body of digital health knowledge.
• To address the limitations of traditional, manual ontology development in the digital health domain.
• To introduce the concept of an "adaptive ontology" for systematic and semi-automatic enhancement of DHT ontologies.

Main Methods:

• Developing a framework that integrates DHT lexicon extraction, ontology enrichment, and human-in-the-loop validation.
• Utilizing large language models (LLMs) to power the adaptive ontology concept.
• Applying the framework to systematically classify and enhance digital health ontologies.

Main Results:

• Demonstration of a semi-automatic approach for classifying and enhancing DHT ontologies.
• Successful implementation of an adaptive ontology powered by LLMs.
• A practical method for managing the complexities of evolving digital health information.

Conclusions:

• The proposed adaptive ontology framework offers an efficient and scalable solution for organizing digital health knowledge.
• LLM-powered adaptive ontologies can systematically enhance the management of DHT information.
• This approach provides a viable path for navigating the dynamic landscape of digital health research and application.