An ontology-supported case-based reasoning approach for damage assessment
Al-Hakam Hamdan, Raimar J. Scherer
Abstract
Despite new technologies in machine vision allow for an automated damage detection, current practices in damage assessment rely mainly on manual evaluations by human experts. Although some new approaches propose a damage assessment via machine learning methods, essential contextual information about the damaged construction is not considered. Contrary to this, knowledge-based approaches have been researched. However, knowledge bases for damage assessment usually contain certain knowledge gaps that result in uncertainties, which still need to be solved manually by experts. Therefore, in this paper a new theoretical approach that utilizes case-based reasoning (CBR) is discussed as additional method for automated damage assessment, which could be utilized in conjunction with knowledge-based approaches. Thereby, the case base of the CBR system would be developed as ontology utilizing the Web Ontology Language (OWL) in order to be compatible with current knowledge-based approaches, especially the Damage Topology Ontology (DOT)