A Comprehensive Methodology to Implement Business Intelligence and Analytics

Abstract: Business intelligence is used by companies for analysing business infor-mation, providing not only historical or current views on business operations, but also providing predictions about the business. Consequently, knowledge discovery in databases can support the implementation of business intelli-gence solutions, especially in order to deal with the reality of big data, using diverse data mining techniques that can help to better prepare the data and to create improved models. The current paper proposes a methodology to im-plement business intelligence and analytics solutions, based on the CRISP-DM methodology, where the application of simplification and equivalence algorithms in modelling data representations can be used for improving the process of business management. This promising approach can boost busi-ness intelligence and analytics by using alternative techniques for discover-ing and presenting new knowledge about the business. The application of simplification and equivalence algorithms within the business context ena-bles finding the most comprehensive or relevant knowledge, represented for instance as association rules, and bringing a real competitive advantage for the stakeholders. . Therefore, these techniques for representing and exchanging learning objects facilitate their sharing among different stakeholders.

Keywords: Business Intelligence, Knowledge Discovery in Databases, Data Mining, Equivalence Algorithm, Canonical Representation.