Aplicando técnicas de Business Intelligence sobre dados das bibliotecas da UNIPAMPA

Abstract

Currently, organizations generate and store large amounts of data related to the activities they perform. However, they have small amounts of available information and therefore, small amounts of available knowledge. This occurs because often the large amount of data makes its analysis and processing impossible by conventional techniques. In this context, emerges Business Intelligence: a set of decision support technologies that provides to managers, executives and analysts a better and faster knowledge discovery in large data sets. Although it has emerged in the business environment, its use does not only apply to this context, but also to support decisions in several domains. Educational institutions, like any other type of organization, aim to provide quality services and academic libraries play a fundamental role in this process because aim support teaching, extension and research. Thus, it was possible to identify the application potential of Business Intelligence in the Federal University of Pampa libraries. The present work presents the main solution steps, which aim to optimize the management of UNIPAMPA libraries through the development of a business intelligence system capable of providing relevant knowledge to the activities developed for libraries and their managers. This work began with the requirements survey and bibliographical research of concepts related to interest of interest, so as to support the design of Data Warehouse (DW) and Data Marts (DMs). The method of extraction, transformation and loading (ETL) and the use of data mining techniques. During ETL, a Talend Data Integration tool was used to extract data from UNIPAMPA databases and upload them to the temporary DW scheme. Also, triggers and functions have been created with a PL/pgSQL language and are responsible for transforming the data and loading it into the main DW schema. Based on the Key Performance Indicators (KPIs), the data sets were given, and with this, it was possible to explore and present the data. The results show that BI is efficient without support to extract useful information, in the form of graphs and information, and the discovery of knowledge capable of supporting the decision process and, consequently, to help the institution management.