Business intelligence in decision-support on communications: Describing a case

• Main Author: Mendes, Armando B.
• Publication Date: 2018
• Other Authors: Alfaro, Paulo Jorge, Ferreira, Aires
• Format: Article
• Language: por
• Source: Repositórios Científicos de Acesso Aberto de Portugal (RCAAP)
• Download full: https://doi.org/10.34627/rcc.v3i0.70

This project addresses decisions of investment on communication infrastructures in Electricidade dos Açores (EDA), the local Electric Company in the Azores Islands. The main decision was that EDA communications should be moving to Voice over IP (VoIP) from present telephone lines, outsourced to an external communications company. At the beginning, a business intelligence project was set, with the objective of getting data from the communications company and analyzing it in order to offer useful information to decision makers. The system uses Microsoft SQL server technologies to establish an OnLine Analytical Processing (OLAP) application. It translates big CSV flat files in a ROLAP infrastructure and presents the results as multidimensional data cubes. Latter some data mining models were implemented and both techniques were incorporated in the CRISP-DM process model. Different models identified several inefficient procedures and even fraud situations as long as supporting the investment decision. These models as long as all the technology developed for gathering data, maintain an manage the OLAP cubes and data mining models were made available to control and strategic decision makers.