Visualizing Information in Data Warehouse Reports

Abstract. The display of information extracted from a data warehouse is an important aspect of human-machine interaction. Visualization tools play a central role in contexts where information must be represented by preserving both the accuracy of data, and the complexity of relationships between data. Much attention has been paid to the problem of effective visualization of data in individual reports that usually are viewed through different types of standard graph (Histogram, Pie, etc.) or in a tabular form. However, this kind of representation provides separate information items, but gives no support to visualize their relationships, which are the basis for most decision processes. This paper presents a methodology for information visualization by exploiting a visual language, called CoDe (Complexity Design), which allows users to manage the complexity of information to represent. In particular, CoDe provides an extensible set of graphical patterns, which allows to compose basic patterns to form complex ones; and allows the designer to specify the semantics of graphs by creating a logical link with the data they represent. The language visually represents both the items of information and their interrelationships at different levels of abstraction, keeping consistency between visually displayed items and the information contained in reports extracted from a data warehouse in tabular form by using the OLAP operations.