Un langage OLAP pour la manipulation d'entrepôts de documents XML

Abstract

OLAP systems based on data warehouses are nowadays well integrated into organizations and they ease the process of information analysis for decision-making. The Web expansion led to increase the volume of the data handled, as well as the sources diversification. This problem was partially solved thanks to the XML language, which indeed allows processing and exchanging complex and heterogeneous data. Only, this format does not fit for classic OLAP systems and warehouse, furthermore there is not an existing standard allowing solving this issue. So we developed a multidimensional model which uses the object oriented model UML to describe a XML Document Warehouse (XDW). The warehouse schema (named StarCD) represents the documents structure in the source, such as it is known by the decision-maker. And we present in this article a new OLAP analysis language intended for decision-makers, which allows them to express complex query on a XDW described by its StarCD.