Challenges in Ontology Competency Question Verification Identified

Ontology engineering relies on Competency Questions (CQs) to ensure an ontology accurately models its intended purpose. However, the process of verifying these CQs is frequently hampered by the intricate interpretation of natural language and the precise mapping required to formal ontology constructs. This complexity can lead to inconsistencies in modeling decisions and verification outcomes. Researchers conducted an experiment with 19 participants using an LLM assistant for CQ verification across 20 tasks. The findings underscore that ambiguities and excessive complexity within CQs significantly complicate the verification process. The study concludes that a crucial step in the ontology engineering workflow should be the refinement of CQs before their publication. This proactive measure would mitigate issues arising from unclear or overly complex questions in later stages, thereby improving the efficiency and accuracy of ontology development.