Rule-based metadata interoperation in heterogeneous digital libraries

摘要

Purpose - The purpose of this research is to describe a system to support querying across distributed digital libraries created in heterogeneous metadata schemas, without requiring the availability of a global schema. Design/methodology/approach - The advantages and weaknesses of ontology based applications were investigated and have justified the utility of inferential rules in expressing complex relations between metadata terms in different metadata schemas. A process for combining ontologies and rules for specifying complex relations between metadata schemas were designed. The process was collapsed into a set of working phases and provides examples to illustrate how to interrelate two similar bibliographic ontology fragments for further query reformulation. Findings - Equipping ontologies with inferencing power can help describe more complex relations between metadata terms. This approach is critical for properly interpreting queries from one ontology to another. Research limitations/implications - A prototype system was built based on examples instead of practical experience. Practical implications - The approach assumes that relations between metadata sets, or ontologies in the approach, are provided by domain experts with/without ontology tools. Originality/value - A new approach has been proposed for facilitating heterogeneous metadata interoperation in digital libraries as a way of empowering ontologies with rich reasoning capabilities. The traditional approach assumes a global schema controlled by a central or virtual server to provide mapping between local and external metadata schemas. A more flexible and dynamic environment was studied, i.e. P2P-based digital libraries, where peers may join and leave freely.