Authors: Daniel Rocha; Orlando Belo
Addresses: ALGORITMI R&D Centre, Department of Informatics, University of Minho, Campus de Gualtar, 4710-057 Braga, Portugal ' ALGORITMI R&D Centre, Department of Informatics, University of Minho, Campus de Gualtar, 4710-057 Braga, Portugal
Abstract: One of the best ways to make an effective selection of data cube views is based on monitoring multidimensional queries during a relevant number of OLAP sessions. This allows to understand how and when a data cube is explored, collecting and analysing the views that decision agents use to consult. This is very important, because it deals directly with the optimisation of resources, namely the ones related to storage capacity and query processing time. Based on this, in this paper, we propose a new view selection method - M3 - for cubes, based on the analysis of OLAP usage sessions. M3 operates on specialised information collected from multidimensional queries launched over one or more data cubes. The aim was to categorise OLAP usage and ensure that views to be materialised will be the ones corresponding to the most widely used and consulted by decision-makers, for a specific period of time.
Keywords: online analytical processing; data warehousing; data cubes; cube view selection; OLAP usage sessions; Markov chains; usage analysis; resource optimisation; storage capacity; query processing time.
International Journal of Decision Support Systems, 2015 Vol.1 No.2, pp.228 - 246
Received: 11 Jul 2013
Accepted: 06 Jul 2014
Published online: 17 Feb 2015 *