Assessing data mining as an alternative in discovering customer consumption pattern / Abdul Malik Hafiy Mohd Jeffri, Muhammad Aqil Zulkifli and Sukri Nasruddin

This study aims to discover customer consumption pattern by using market basket analysis. In discovering the pattern, the researcher use McDonald's restaurant franchise located at Seremban 2. This study was carried out using primary data. The data had been collected by the transaction done by customer in form of receipt. Data were analysed by using Rstudio software and apriori algorithm is used for obtaining the pattern. The first objective is to generate the frequent item sets that are frequently bought by customers. The findings of the study show that most of the item sets are scattered between support value of 0 to 0.05. The second objective is to generate the association rule between frequent item sets. The findings also show that 137 rules have been generated after removing redundant rule with lift value ranging from 2 to 50. The third objective is to discuss the most interesting rule that is generated by apriori algoritm. The findings show customer that buy main meal set need more time in deciding drinks whether to choose default drinks or change to other drinks. This means that throughout daily activities, customer preferences changes over time. There are many different pattern of customer preference being discovered through the analysis.