Vertical Association Rule Mining: Case study implementation with relational DBMS

Data mining remains as one of the most important research domain in Knowledge Discovery and Database (KDD). Moving deeper, Association Rule Mining (ARM) is one of the most prominent areas in detecting pattern analysis especially for crucial business decision making. It aims to extract interesting correlations, frequent patterns, association or casual structures among set of items in the transaction databases or other data repositories. Even though most of these data repositories are dealing with flat files, current trend is focusing on using relational Database Management System (DBMS) for the more systematic and structured management of data. In response to the importance of adopting relational database, in this paper we implement MySQL (My Structured Query Language) as our association rule mining database engine in testing benchmark dense datasets which available from Frequent Itemset Mining (FIMI) online repository. Our study is focusing on Eclat algorithm as well as its variants in generating frequent and interesting rule, as a continual from our previous studies. The performance result shows a promising signal as to confirm on the benefits of relational database mechanism in storing any transaction data.