Market Basket Analysis: A Profit-Based Product Promotion Forecasting

Abstract

mining is the area that helping extracting the useful information by finding patterns or rules from the existing dataset. By using the extracted information then used to predict future tendencies and behavior patterns. Association mining is a branch of data mining which used to identify itemsets that take place frequently in a specific dataset and to determine rules. Association mining can find out the rules that predict the occurrence of an item with regard to the similar occurrences of other in a particular transaction. Eclat algorithm is kind of a frequent itemset mining which is a sub section of the association mining based on the mining frequent patterns by exploring the vertical data format. Eclat algorithm was actually developed for Market Basket Analysis which is an effective technique in retail industry that helps the shop owner to increase the sales distribution techniques. Market Basket Analysis is completely done by the association rule mining in which analyses the customer buying behavior against the purchasing item from the shop. Eclat algorithm is the one of the most effective ways to mining of large data set since it follows the depth in search. When it comes to the real world, the main objective of market basket analysis is to gain maximized profit at all with the help of operational research theories. In this approach, the condensed data is used for mine the frequent itemset using the Eclat algorithm. After all, one of the operational research theories which are termed linear programming will use to maximize the profits. Support value and the Confidence value are the foremost factors in generating the Eclat. Eclat algorithm abandons Apriori’s breadth-first search for a recursive depth-first search. Moreover, consideration of frequent items as well as non-frequent items, considerably impact the profit maximization. Because if the retail owner identified the non-frequent itemset; can provide the promotions to the customers. It will enhance the profit maximization. Therefore, this research was mainly focused to identify frequent itemset as well as the non-frequent itemset in a market basket analysis alone with the profit maximization using linear programming. This developed approach is applied to a real world dataset and results were compared considering Eclat algorithm and Eclat algorithm alone with the linear programming separately. Finally, the results conclude that proposed approach significantly increase the profit.