Malaysian Journal of Computer Science (ISSN 0127-9084)
Indexing Page
Visit the official web site at http://ejum.fsktm.um.edu.my

Article Information
Title:Student Academic Streaming Using Clustering Technique
Auhtor(s): Ely Salwana,Suraya Hamid,Norizan Mohd Yasin,
Journal:Malaysian Journal of Computer Science (ISSN 0127-9084)
Volume:30, No 4
Year:2017
Keywords:studentís performance; data mining; clustering; academic streaming
Abstract:The balance of human capital supply and industry demands are crucial to sustain a competitive advantage in order to ensure stability of economic growth. Unfortunately, companies often find it hard to recruit the right people. Many ideas obviously relate to connect human capital to education because human capital is created through education. In the education sector, there is a critical need to develop an effective planning mechanism to distribute the students into the most suitable area in the industry. In order to achieve this, the studentís pathway needs to be planned systematically in school by identifying the studentís streaming based on their academic performance. In this case, students who have the same performance will be grouped in the same cluster using a data mining technique. However, the problem is that, it is difficult to identify real potential for the students, because their performance is not well monitored, and current assessment systems do not support the studentís academic planning activities. Besides, there is no specific technique used to group students into clusters based on their performance. This study aims to overcome the problem by grouping the students according to their performance using clustering techniques and to propose a suitable model. This study aims to overcome the problem by identifying a suitable clustering model that can be used to analysis an educational data. The data involves is student performance data. Based on the data, two clusters of students are created which is science and arts. A novelty in the method of study is the use of three clustering models and a comparison among them in order to find a suitable clustering model to be used with student academic performance data. The study was conducted in five schools in Malaysia to support studentsí grouping in two different academic streams, which are science and art. The result demonstrated the best model of clustering technique that is suitable for mining the educational data. Moreover, suitable streaming based on the studentsí performance and education policy was created from the results. It can be used to assist schools and students in determining the appropriate streaming for the students, and support for the human capital needs by the country in the future.
File:Download

Volume Listing