Storing and querying XML (eXtensible Markup Language) data in relational form can exploit various services offered by modern relational database management systems (RDBMSs). Due to structural complexity of XML, there ...Storing and querying XML (eXtensible Markup Language) data in relational form can exploit various services offered by modern relational database management systems (RDBMSs). Due to structural complexity of XML, there are many equivalent relational mapping schemes for the same XML data and queries. In this paper, we propose the adaptive XML to relational mapping (AX2RM) system, which considers finding optimal XML to relational (X2R) mapping as four separate but correlated procedures: logical database design, data scale estimation, workload transformation, and physical database design. We view the whole process as an autonomic computing problem and formalize the adaptive X2R mapping problem. Search spaces for each procedure are investigated individually, and five approaches for finding the optimal mapping are studied. We propose an integrated approach with greedy pruning (IT-GP), which views the mapping procedures as a whole and exploits heuristic rules in each procedure to prune impossible mappings as early as possible. Evaluation of these approaches shows the validity and high efficiency of IT-GP.展开更多
基金the National Natural Science Foundation of China (No. 60603044)the China Postdoctoral Science Foundation (No. 20070411179)the Program for Changjiang Scholars and Innovative Research Team in University of China (No. IRT0652)
文摘Storing and querying XML (eXtensible Markup Language) data in relational form can exploit various services offered by modern relational database management systems (RDBMSs). Due to structural complexity of XML, there are many equivalent relational mapping schemes for the same XML data and queries. In this paper, we propose the adaptive XML to relational mapping (AX2RM) system, which considers finding optimal XML to relational (X2R) mapping as four separate but correlated procedures: logical database design, data scale estimation, workload transformation, and physical database design. We view the whole process as an autonomic computing problem and formalize the adaptive X2R mapping problem. Search spaces for each procedure are investigated individually, and five approaches for finding the optimal mapping are studied. We propose an integrated approach with greedy pruning (IT-GP), which views the mapping procedures as a whole and exploits heuristic rules in each procedure to prune impossible mappings as early as possible. Evaluation of these approaches shows the validity and high efficiency of IT-GP.
