Patterns Identification in the Consumption of Oracle Database Resources Using Weka Tool in Data Mining

Abstract

This article focuses on the identification of periods of consumption of database resources helping the management of I.T. in decision-making. The purpose of this work is to mine data dictionary information from a given database, the Oracle Database System Manager, to obtain information to assist in analyzing the features of this database. For this, the KDD information extraction process was used in order to extract knowledge from the database by the identification of patterns. As a tool, we used Weka to develop this study, because it is open-source and has a range of algorithms that contains mining techniques, where it is possible to apply the last two steps of KDD. The applied methodology was a case study, making use of bibliographical techniques to collect information pertinent to the topic, in which all the 5 attributes related to the computational resources utilization of an Oracle DBMS were used. This data set consists of 5 attributes and 120 instances. Analyzing the statements generated by the rules it is possible to know which resource consumption is more or less used during a period of the day, proving to be useful in determining the resource consumption characteristics and thus assist in the decision to define a specific time to execute processes without affecting the overall performance of the database.