DATA MINING APPROACHES IN BUSINESS INTELLIGENCE: POSTGRADUATE DATA ANALYTIC

Abstract

Over recent years, there has been tremendous growth of interest in business intelligence (BI) for higher education. BI analysis solutions are operated to extract useful information from a multi-dimensional datasets. However, higher education-based business intelligence is complex to build, maintain and it faces the knowledge constraints. Therefore, data mining techniques provide an effective computational methods for higher education-based business intelligence. The main purpose of using data mining approaches in business intelligence is to provide decision making solution to higher education management. This paper presents the implementation of data mining approaches in business intelligence using a total of 13508 postgraduates (PG) data. These PG data are to allow the research to identify the postgraduates who Graduate On Time (GOT) via business intelligence process integrating data mining approaches. There are four layers will be discussed in this paper: data source layer (Layer 1), data integration layer (Layer 2), logic layer (Layer 3), and reporting layer (Layer 4). The main scope of this paper is to identify suitable data mining which is to allow decision making on GOT so as to an appropriate analysis to education management on GOT. The results show that Support Vector Machine (SVM) classifier is with better accuracy of 99%. Hence, the contribution of data mining in business intelligence allows an accurate decision making in higher education.