Towards the automation of XML data warehouse logical design

Over the past few years XML has rapidly become a widely accepted data format for information interchange and representation of semi-structured data. As a result, huge amounts of interesting data are generated and distributed on the Web, prompting the need of new and automated approaches to store, organize and analyze such data. Data warehouses systems provide an efficient tool and integrated repository with a high level of consolidation and a multi-dimensional view of the data. Several studies address the issue for designing a conceptual and logical schema of XML Data Warehouses. Nevertheless, these approaches focus only on identifying multidimensional concepts in a semi-automatic fashion. In this paper we propose an automatic approach for designing the logical schema for a data mart starting from the XML schema describing XML sources using UML and QVT transformation language. Our focus is not only on automating an XML Data Warehouse design, but also on providing a simplification process and a set of rules that applies successive transformations to create the star schema which is the predominant logical multidimensional scheme. Finally, we discuss implementation issues and we present a graphical tool to help designers and engineers design without technical expertise in multi-dimensional data modeling.