Data warehousing tool's architecture: from multidimensional analysis to data mining

Decision support tools evolve quickly. End user needs for analyses are more and more sophisticated. That is why it is necessary to set up departmental data marts from the data warehouse. Usually these data marts use the OLAP architecture, offering a multidimensional view of data. There are four main categories of analysis tools: query tools, multidimensional data bases (MOLAP), relational OLAP (ROLAP) and data mining. Query tools are light analysis tools. Multidimensional databases are databases stored in array table form which allow rapid visual representation of basic consolidated data from multiple analysis viewpoints. Relational OLAP offers the same capabilities as OLAP but doesn't store any multidimensional database and interrogates the relational database directly. Among forthcoming evolutions in the decision support domain, data mining is certainly the most important. This paper presents the different data warehousing technologies.