Authors: Rahma Djiroun; Kamel Boukhalfa; Zaia Alimazighi
Addresses: LSI Laboratory, USTHB University, Algiers, Algeria ' LSI Laboratory, USTHB University, Algiers, Algeria ' LSI Laboratory, USTHB University, Algiers, Algeria
Abstract: Business intelligence systems provide an effective solution from large volumes of data for multidimensional online computing and analysis. Usually, in a decision-making process, organisations and enterprises, require several internal and/or external cubes which are often heterogeneous. Most of the time, the structure of these cubes is unknown to the decision-makers. To analyse a phenomenon, the decision-maker seeks among sets of cubes, in a collection, the cube which responds better to his need. In this context, we propose an approach that enables decision-makers to express their needs via a query expressed in a natural language, returns top-K relevant cubes and designs/constructs new cubes when no, or few deployed cubes are relevant. We propose a tool called RD-cubes-query implementing our approach in a ROLAP architecture. We use this tool in some experiments to validate our approach.
Keywords: cubes design; cubes search; online analytical processing; OLAP; top-K; query analysis; visualisation tool.
International Journal of Business Intelligence and Data Mining, 2019 Vol.14 No.1/2, pp.267 - 298
Received: 31 Jul 2017
Accepted: 19 Nov 2017
Published online: 16 Nov 2018 *