Bootstrapping Ontology Alignment Methods with APFEL

Abstract. Ontology alignment is a prerequisite in order to allow for interoperation between different ontologies and many alignment strategies have been proposed to facilitate the alignment task by (semi-)automatic means. Due to the complexity of the alignment task, manually defined methods for (semi-)automatic alignment rarely constitute an optimal configuration of substrategies from which they have been built. In fact, scrutinizing current ontology alignment methods, one may recognize that most are not optimized for given ontologies. Some few include machine learning for automating the task, but their optimization by machine learning means is mostly restricted to the extensional definition of ontology concepts. With APFEL (Alignment Process Feature Estimation and Learning) we present a machine learning approach that explores the user validation of initial alignments for optimizing alignment methods. The methods are based on extensional and intensional ontology definitions. Core t...

• Type: Conference
• Year: 2005
• Where: SEMWEB
• Authors: Marc Ehrig, Steffen Staab, York Sure
• Keywords: Alignment Task, Machine Learning, Ontology Alignment, SEMWEB 2005