Identifying Granularity Differences between Large Biomedical Ontologies through Rules.

The objective of this study is to identify the granularity differences as well as similarity between large biomedical ontologies through rules. Two anatomical ontologies were selected, and based on a set of concept mappings obtained through simple string matching techniques, we constructed rules to distinguish among different types of subclasses and classifications. 82% of the concept mappings have exactly the same classification in subclasses between the two ontologies. Other mappings are classified in different granularity, including additional subclasses, detailed classification, and different intermediate classification concepts. Using rules and the rule inference engine enables an automatic and scalable investigation of the structural incompatibility among biomedical ontologies.