Modulated evaluation metrics for drug-based ontologies.

BACKGROUND

Research for ontology evaluation is scarce. If biomedical ontological datasets and knowledgebases are to be widely used, there needs to be quality control and evaluation for the content and structure of the ontology. This paper introduces how to effectively utilize a semiotic-inspired approach to ontology evaluation, specifically towards drug-related ontologies hosted on the National Center for Biomedical Ontology BioPortal.

RESULTS

Using the semiotic-based evaluation framework for drug-based ontologies, we adjusted the quality metrics based on the semiotic features of drug ontologies. Then, we compared the quality scores before and after tailoring. The scores revealed a more precise measurement and a closer distribution compared to the before-tailoring.

CONCLUSION

The results of this study reveal that a tailored semiotic evaluation produced a more meaningful and accurate assessment of drug-based ontologies, lending to the possible usefulness of semiotics in ontology evaluation.