1 Mining Least Association Rules of Degree Level Programs Selected by Students Tutut Herawan 1 , Zailani Abdullah 2 , Wan Maseri Wan Mohd 1 , A Noraziah 1 1 Faculty of Computer Systems and Software Engineering Universiti Malaysia Pahang, Malaysia Lebuhraya Tun Razak, 26300 Kuantan Pahang, Malaysia 2 Department of Computer Science Universiti Malaysia Terengganu 21030 Kuala Terengganu, Terengganu, Malaysia {tutut,maseri,noraziah}@ump.edu.my, zailania@umt.edu.my Abstract. One of the most popular and important studies in data mining is association rules mining. Generally, association rules can be divided into two categories called frequent and least. However, finding the least association rules is more complex and time consuming as compared to the frequent one. These rules are very useful in certain application domain such as determining the exceptional association between university’s programs being selected by students. Therefore in this paper, we apply our novel measure called Definite Factors (DF) to determine the significant least association rules from undergraduate’s program selection database. The dataset of computer science student for July 2008/2009 intake from Universiti Malaysia Terengganu was employed in the experiment. The result shows that our measurement can mine these rules and it is at par with the existing benchmarked Relative Support Apriori (RSA) measurement. Keywords: Data Mining; Association rules; Significant Least, Measure; Educational Data. 1 Introduction Data mining can be defined as the process of extracting hidden and useful information from large data repositories [1]. One of the emerging interdisciplinary research areas in data mining is educational data mining [2]. By definition, educational data mining is an application of suitable data mining techniques to analyze the educational data [3]. It aims at developing new methods that can discover the interesting information from educational settings, and used those methods to better understand the students, and their learning settings (http://www.educationaldatamining.org). The problem of association rules mining was first coined by [4] in an attempt for market-basket analysis. The classification of frequent or least items is based on the mechanism of support threshold. A set of items (itemset) is said to be frequent, if it appears more than minimum support count. The item (or itemset) support count is defined as a probability of item (or itemset) appears in the transaction. In addition, confidence is