Descoberta de conhecimento sobre o processo seletivo da ufla

Abstract

The information has been having a fundamental role on companies' growth, development and sucess. The making-decision supporting systems, available at these companies, make the work of collecting, treating and analyzing. There is also a tendency in these companies to increase their data amount. However, there is an inverse relation between the data amount and the need of a strategic knowledge, that is, although the resumed and meaningful information to making- decision are fewer, generally they are not available and demand to be extracted from big data amounts. KDD - Knowledge Discovery in Databases refers to the extration of knowledge from a big database amounts. Data Mining refers to a especific phase of this process. This study demostrates a practical application of KDD Process to the database of 2006 UFLA's Entrance Examination or Selective Process. Coherently to WEKA research tool - Weikato Enviroment for Knowledge Analysis, the Data Visual Mining, Decision Tree, Association Rules and Neural Networks were applied. The results can be used to biuld up the candidates profile, in order to extract important information that offer support to this institution on the making-decision process.