University of Konstanz
Graduiertenkolleg / PhD Program
Computer and Information Science

Graduation Talks


Extending the OLAP technology to support non-conventional application domains


Svetlana Mansmann, University Konstanz
Konstanz, Germany

date & place

Wednesday, 23.05.2007, 15:15 h
Room C252


Comprehensive data analysis has become indispensable in a variety of real-world applications. However, the business-oriented OLAP (On-Lin eAnalytical Processing) systems tend to perform poorly or even fail when applied to complex data scenarios found in novel application domains, such as government, science, medicine etc. The causes of the deficiencies are manyfold, from the rigidness of the underlying multidimensional data model, the conventional logical schemata, and the data transformation techniques to the established metadata modeling standards and end-user frontend issues, such as visual navigation and query specification, available visualization and interaction techniques, and user friendliness.The presented research is concerned with overcoming the above mentioned limitations of the relational OLAP systems by analyzing the requirements of non-conventional application scenarios towards such systems and proposing the modeling extensions at the conceptual, logical and metadata levels as well as their implementation in a visual exploration framework. The proposed modeling extensions are captured accordingly at the metadata level and can be handled using the standard database query language SQL. The major front-end contributions, namely a powerful context-aware schema-based navigation for multidimansional data cubes and a class of hierarchical visualization techniques called Enhanced Decomposition Trees, are presented. The feasibility of the proposed enhancements is demonstrated using realistic study cases from the fields of academic administration and medical research.