RDF-based Metadata Management in Peer-to-Peer Systems
Huiyong Xiao, Isabel F. Cruz · 2004
Abstract. Peer-to-Peer (P2P) data integration combines the P2P infrastructure with traditional schema-based data integration techniques. One of the primary problems in this research area is the automation of the process of interoperation among heterogeneous data sources. To facilitate this process, metadata are used to add rich semantics to the data sources. In this paper, we propose an RDF-based layered approach to data interoperation in a P2P environment. We use RDF and its vocabulary language RDFS for representation of data and metadata, which are distributed in different peers. For the representation of the mappings of metadata from peer to peer (called P2P mappings), we propose a mapping language, which is an RDF-based meta-ontology (called RDFMS) that enables metadata management in a decentralized way. We also discuss a thesaurus-based mapping process and query translation in the proposed framework. 1