Current Proceedings on Technology

Yazarlar: Zailani Abdullah, Md Yazid Mohd Saman, Tutut Herawan, A. Noraziah, Mustafa Mat Deris

Konular:-

Anahtar Kelimeler:Interesting association rules,Educational,Correlation.

Özet: One of the popular techniques in data mining is association rules mining. It aims at discovering interesting relationship among set of item in transactional databases. Association rules are very useful in main domain applications in term of correlation analysis and decision making process. In educational context, determination a “right” program for student if their chosen programs are not selected is quite challenge. Typically, the process of assigning the programs for that category of students is based on the minimum qualifications and availability of the respective programs. The student interest which can be captured from the history of their chosen program is not taken into account. As a result, they might have a problem in term of academic commitment and academic achievement. Therefore, Therefore in this paper, we proposed a model which consists of pre-processing, mining patterns and assigning weight to discover highly positive association rules. We examined the previous chosen programs by computer science students in our university for July 2008/2009 intake. The result shows that the proposed model can mine association rules with high correlation. Moreover, for data analysis, there are existed students that have been offered in computer science program at our university but not within their program interests.

---