Benchmarking LLM-based ontology conceptualization: A proposal