Schema Versioning in Data Warehouses
The deployment of data warehouses, which are integrated databases to support decision making, has become current practice in modern information technology landscapes, and the methodical design process for such databases has been advanced significantly in recent years. In this context we are investigating questions concerning the automatic design and quality of data warehouse schemata.
We proposed multidimensional normal forms to measure and algorithmically enforce schema quality during conceptual design. These normal forms formalize correctness and completeness of data warehouse schemata with respect to the underlying operational databases, avoidance of redundancies, and context sensitive summarizability in the presence of null values. Our tool Data Warehouse Detective enables the design of data warehouse schemata based on multidimensional normal forms.
In view of complex dimension hierarchies we are investigating rich conceptual modeling approaches, their impact on summarizability, and automatic transformation approaches to implement rich conceptual models based on standard relational database technology.
Contact
Street:Leonardo-Campus 3
City:48149 Münster
Country:Germany
Web:Website
Email: vossen@ercis.uni-muenster.de
News
[Tue, 01/31/2012]
In the course of the final presentation of the joint ERCIS seminar “Global Challenges”, Dr. Bernd Schenk, representing the University of Liechtenstein, and Dr. Armin Stein, representing the...
[Mon, 01/16/2012]
During The 24th International Conference on Advanced Information Systems Engineering (CAiSE'12) in Gdańsk on June 25th till 29th, 2012
...
[Mon, 01/09/2012]
Stefan Stieglitz and Linh Dang-Xuan were awared with the Best Paper Award for their publication "Political Communication and Influence through Microblogging – An Empirical Analysis of...
[Mon, 12/05/2011]
ECIS 2013 will be held in Utrecht, the Netherlands. The organizers have opened the call for Track Proposals and as a program co-chair our partner Jos van Hillegersberg is looking for innovative...
