BBT: An Efficient Association Rules Mining Algorithm using Binary-Based Technique

Association rules mining is one of the important tasks in data mining and knowledge discovery in database. It requires the discovery of frequent patterns from datasets before any association between the items can be extracted. Finding frequent itemsets is computationally expensive and I/O intensive especially with the growing size of today's database datasets. Although numerous works have been carried out to handle the above issues, there still exists the need for more explorations and developments in that area. In this paper, an efficient Binary-based algorithm for mining association rules in large-scale databases is introduced. The proposed approach is also able to minimize the execution time and main memory usage, when compared to some of the well-known algorithms. Extensive experiments have been carried out to validate the technique; and based on the results obtained, it outperforms Apriori, Eclat, FPgrowth, and H-mine algorithms in terms of execution time and memory usage.