Ontologies are a common tool for representing complex knowledge in fields like biomedicine and healthcare. But constructing reliable ontologies is difficult, especially when class hierarchies are built manually or through other existing methods.
A new paper on arXiv investigates how embeddings and large language models (LLMs) can help with ontology refinement. The authors use CASTON as a case study to test this approach.
Because the abstract is brief, the specific methods and results are not detailed here. The work points to a growing interest in using modern AI techniques to support the maintenance and improvement of structured knowledge resources.