Abstract
Data sharing in large-scale Peer Data Management Systems (PDMS) is challenging due to the excessive number of data sites, their autonomous nature, and the heterogeneity of their schema. Existing PDMS query applications have difficulty to simultaneously achieve high recall rate and scalability. In this chapter, we propose an ontology-based sharing framework to improve the quality of data sharing and querying over large-scale distributed communities. In particular, we add a semantic layer to the PDMSs, which alleviates the semantic heterogeneity and assists the system to adjust its topology so that semantically related data sources can be connected. Moreover, we propose a comprehensive query routing and optimization scheme to enable scalable and efficient query evaluation. Simulation studies reveal that our proposed ontology-based data sharing framework effectively improves the query performance in terms of recall and scalability.