Searching for efficient XML-to-relational mappings

We consider the problem of cost-based strategies to derive efficient relational configurations for XML applications that subscribe to an XML Schema. In particular, we propose a flexible framework for XML schema transformations and show how it can be used to design algorithms to search the space of equivalent relational configurations. We study the impact of the schema transformations and query workload on the search strategies for finding efficient XML-to-relational mappings. In addition, we propose several optimizations to speed up the search process. Our experiments indicate that a judicious choice of transformations and search strategies can lead to relational configurations of substantially higher quality than those recommended by previous approaches.