Description
The objective of this article is to recapitulate our experience in aligning large anatomical ontologies (Foundational Model of Anatomy, GALEN, Adult Mouse Anatomical Dictionary and NCI Thesaurus) having different representation formalisms. Our approach to aligning concepts (directly) is automatic, rule-based, and operates at the schema level, generating mostly point-to-point mappings. It uses a combination of lexical, structural and semantic techniques. It also takes advantage of domain-specific knowledge (lexical knowledge from external resources, such as the Unified Medical Language System, as well as knowledge augmentation and inference techniques). In addition to point-topoint mapping of concepts, we present the alignment of relationships and the mapping of concepts group-to-group. We have also successfully tested an indirect alignment through a reference ontology. We present an evaluation of our techniques, both against a gold standard and against a generic schema matching system. The advantages and limitations of our approach are analyzed and discussed throughout the article.