An Adaptive Implementation Case Study of Apriori Algorithm for a Retail Scenario in a Cloud Environment

Retail transactional databases are voluminous and traditional algorithmic approaches to mine pattern in them are time consuming. The current study presents an approach to scale Apriori a Frequent Item set Mining (FIM) algorithm, which is often used for market basket analysis. The study also compares the performance of scaled version of the algorithm (running on multiple on-demand simultaneous Azure cloud instances) with that of traditional setup (running on a fixed Azure cloud instances) using simulated data sets. The experimental results show that the response times were significantly lower and in favor of the scaled approach as the data volume increases.