Using a factorial approach for efficient representation of relevant OLAP facts

On line analytical processing (OLAP) is a technology basically created to explore data cubes and detect relevant information. Unfortunately, in huge and sparse data volumes, exploration becomes a tedious task. In such a case, simple user's intuition or experience does not always lead to efficient results. In this paper, we propose to exploit the multiple correspondence analysis (MCA) in order to assist exploration of cubes by enhancing their space representations. MCA is a factorial method that maps associations of huge number of categorical variables and displays them within an appropriate space representation. Our approach uses test-values provided by MCA in order to detect and arrange OLAP facts in a large and sparse data cube within an interesting visual effect which gathers full cells in relevant regions and separates them from empty cells. Thus, it is possible to focus analysis on interesting facts by browsing directly the provided regions in the data cube