A flexible biomedical ontology selection tool

Abstract

The wide adoption and reuse of existing biomedical ontologies available in various libraries is limited by the lack of suitable tools with metrics for their evaluation by both naïve users and expert ontologists. Existing evaluation tools perform technical evaluation of the structure of an ontology as presented by its design and knowledge representation. These find little use in evaluating the processes and representation of granularity presented by biomedical ontology. In this paper we present an evaluation tool as part of a flexible framework that enables users to select a suitable biomedical ontology for use in building applications that integrate clinical and biological data. Requirements for such a tool were elicited in a descriptive survey using questionnaires, and a prototype developed. The tool also enables ontology modelers to iteratively elicit new requirements for improving upon existing biomedical ontologies, leading to new ones that are able to integrate data across structure, processes and granularity. By facilitating biomedical ontology evaluation, the tool contributes towards reuse of existing biomedical ontologies. This helps to avoid the need for costly time consuming tasks of developing entirely new biomedical ontologies. The utility of this tool was demonstrated in experiments that evaluated the infectious disease ontology. The results were validated using a questionnaire based human assessment.