Mineração de dados como parte do processo de descoberta do conhecimento

Abstract

The Knowledge Discovery in Databases, or Knowledge Discovery in Databases (KDD) is a process that consists of six stages: data selection, data cleansing, enrichment, transformation or data coding, data mining and reporting and display of information discovery. Data mining is a nontrivial process of extracting or mining knowledge previously unknown, but potentially useful for large volumes of data in order to identify anomalies, establish patterns, relationships or associations that transform raw data into valuable information strategically. It is processed to identify several factors and trends in business activities, quality control, fraud detection, scientific research, process control, among others. Waikato Environment for Knowledge Analysis (WEKA) is a free software like open source for data mining. Was developed in the Java programming language for the course of Computer Science, University of Waikato, New Zealand, in 1993, within the specifications of the GPL (General Public License). Consolidated over the last few years, as the tool most widely used data mining in academic environment. This paper presents the data mining performed in historical students of Bachelor of Information System (BSI) of the Federal University of Grande Gold (UFGD) in the period 2006- 2009, in order to discover the causes or potential key factors that cause high rates of repetition and dropout and low approval ratings and graduation.