Authors: Maurizio Pighin, Lucio Ieronutti
Addresses: IS&SE-Lab, Department of Math and Computer Science, University of Udine, Via delle Scienze 206, 33100 Udine, Italy. ' ISSE-Lab, Department of Math and Computer Science, University of &Udine, Via delle Scienze 206, 33100 Udine, Italy
Abstract: Data warehousing allows analysts and managers to interactively study transactional data for deriving information suitable to support decision processes. However, the effectiveness of such tools strongly depends on taken design choices and quality of initial data. The selection of proper measures and dimensions is not a trivial task, especially in the case of large transactional databases. We propose an approach based on statistical and syntactical aspects that can be effectively used for: guiding the attributes selection during the datawarehouse design; evaluating the quality of taken design choices. We tested on three commercial databases the effectiveness of our proposal.
Keywords: data warehouse design; quantitative quality measurements; data warehouse evaluation; data warehousing; attributes selection.
International Journal of Information Quality, 2007 Vol.1 No.4, pp.368 - 391
Available online: 17 Jan 2008 *Full-text access for editors Access for subscribers Purchase this article Comment on this article