Hidden Knowledge Extraction for Association Rule Mining Technique Based on the Apriori and Frequent Pattern Growth Algorithms: A Case Study

Association rule mining technique is a branch of data mining area. It’s purpose is to extract the hidden knowledge from databases D with transactions T. Due to the huge amount of data from multiple sources continuously being captured and stored in databases D, many industries, enterprises, to name just a few, are more interesting in exploring association rules from their transactional databases in order to increase their profits. In this paper, the simulation study is investigated based on the Apriori and Frequent Pattern Growth (FP-G) algorithms which are popular utilized in the discovery of the relevant hidden frequent items from transactional databases. This study focuses on the design and the implementation of the appropriate data structure using linked lists of array. The simulation of the algorithms is evaluated using a supermarket dataset named as retail market basket data so as to analyze the performance characteristics of the implemented algorithms, on a 8GB Random Access Memory (RAM) laptop computer. Furthermore, the investigation study uses the designed useful data structure and the features of the algorithms during the extraction of frequent itemsets on supermarket dataset, which are then used for generating the association rules, and performance costs estimated for different values of association rule length, minimum support, and time-consuming performance.