Intelligent data cube construction and exploration

Abstract

Data cubes are multi-dimensional structures that consist of dimensions and measures. Analysts can view the performance measures via different perspectives provided by the available dimensions in data cubes. However, modeling of these meaningful dimensions and selection of informative measure is a difficult task for human data warehouse developers. In high dimensional environments, the sheer size and volume of data poses a number of challenges in order to generate meaningful data cubes. Nowadays, there is a growing requirement of automated and intelligent techniques that allows analysts to construct and explore the large cubes for better decision making. In this paper, we have reviewed the literature on intelligent data cubes construction and exploration. Literature review reveals that a number of techniques have been proposed to embed intelligence in data cubes. However, majority of the previously proposed technique targeted either on providing intelligence in cube construction or focused assisting the intelligent exploration of data cubes. However, there is very limited amount of work has been done in the integration of intelligent techniques for both cube construction and exploration. We believe that it is a strong area of research and the modern analytical systems demand the availability of intelligent techniques for both construction and exploration of large data cubes for making intelligent decisions. The objective of this paper is to present a critical review of the existing techniques and to propose a conceptual model that not only overcomes the individual limitations in the previous work but also merges the benefits of intelligent construction and exploration of cubes in parallel. However, the implementation of the proposed model is beyond the scope of this paper.