Mining Interesting Least Association Rules in Manufacturing Industry: A Case Study in MODENAS
Yunus Indra Purnama, Zailani Abdullah, Rokhmat Rokhmat, Tutut Herawan · International Journal of Control and Automation · 2015
Least association rules are related to the rarity or uncommonness relationship among itemset in database repository. However, mining these rules are quite tricky and seldom discussed since it usually includes with infrequent items or exceptional cases. In manufacturing industry, detecting these rules is very useful and exciting for further analysis such as for market segmentation, prediction, and product arrangements. In this paper, we introduce an enhanced association rules mining method, called Significant Least Pattern Growth (SLP-Growth) and a new measurement named Critical Relative Support (CRS). The novelty of the proposed method is that unlike existing methods, it is used for capturing interesting least items from line-off database. The data which comprised of production of motorcycle/scooter is taken from Malaysia national motorcycle manufacturer called Motorsikal dan Enjin Nasional Sdn Bhd (MODENAS). The results from this research provide useful information for the management to understand the customer demand trends comprehensibly, and enable them to design marketing strategy accordingly.