An approach to query relaxation using ontologies in a GIS-based archiving system

The MonArch Digital Archiving System is a metadata repository designed for the storage, management and retrieval of digital documents and digital information in general. In contrast to conventional document management systems, the MonArch System uses semantical technologies to improve content descriptions. Furthermore, GIS functionality and navigation maps allow for the geo-referencing of documents and thus for visually locating and selecting the archived content. Together, the semantical and spatial markups allow for powerful queries that are fully adaptable to the needs of a wide variety of scenarios. In a federated network of archives each repository could benefit from the accumulated but distributed information. To capitalise on this advantage, coping with the inevitable disparities of the description schemata is one of the major challenges. Therefore, a distributed query facility must be able to bridge different schemata as well as the particular domain and purpose of each repository. In this paper an approach to query relaxation in the application domain of digital preservation of built heritage is presented. The proposed method is based on similarity which in turn can be inferred from the structural properties and a type-based classification of the descriptive metadata attached to the stored objects. The resulting approach reduces significantly the otherwise necessary manual schema integration work which tends to rely on domain experts.