Authors:
Maurizio Pighin
and
Lucio Ieronutti
Affiliation:
University of Udine, Italy
Keyword(s):
Datawarehouse, design quality, data quality.
Related
Ontology
Subjects/Areas/Topics:
Data Warehouses and OLAP
;
Databases and Information Systems Integration
;
Enterprise Information Systems
Abstract:
Data warehousing provides tools and techniques for collecting, integrating and storing a large number of transactional data extracted from operational databases, with the aim of deriving accurate management information that can be effectively used for supporting decision processes. However, the choice of which attributes have to be considered as dimensions and which as measures heavily influences the effectiveness of a data warehouse. Since this is not a trivial task, especially for databases characterized by a large number of tables and attributes, an expert is often required for correctly selecting the most suitable attributes and assigning them the correct roles. In this paper, we propose a methodology based on the analysis of statistical and syntactical aspects that can be effectively used (i) during the data warehouse design process for supporting the selection of database tables and attributes, and (ii) then for evaluating the quality of data warehouse design choices. We also p
resent the results of an experiment demonstrating the effectiveness of our methodology.
(More)