Ontology learning in complex domains, such as life sciences, poses significant challenges for current Large Language Models (LLMs). Existing models struggle with generating deep, structured ontologies due to token limitations, shallow hierarchies, and inadequate domain adaptation. To address these issues, we extend the NeOn-GPT pipeline with advanced prompt engineering techniques and ontology reuse to enhance domain-specific reasoning and structural depth of the generated ontologies. Our work evaluates the capabilities of LLMs in ontology learning in the context of life science domains, and we use the AquaDiva ontology as a case study to assess the logical consistency, completeness, and scalability of the generated ontologies. Our evaluation shows the viability of LLMs for ontology learning in specialized domains, providing solutions to longstanding limitations in model performance and scalability.
- File Directory:
Overview of our proposed methodology to extend the NeOn-GPT pipeline for ontology learning to suite more complicated domains such as life science domains.
