Short Time Comparison of Item Sets With Market Basket Analysis Using Frequent Pattern (FP) Growth Method

Abstract

Supermarket analysis examines purchasing patterns to identify relationships among the various goods in a customer's shopping cart. The results of these correlations have assisted businesses in developing a successful sales strategy by grouping goods that customers commonly buy. Due to the expanding volume of data and its widespread utilization in the retail sector to enhance marketing strategies, Data Mining (DM) has become increasingly important in recent years. Transaction data analysis from the past yields a wealth of knowledge about consumer behaviour and commercial choices. The rate at which data is saved doubles every second that the fastest processor is available for its analysis. In a huge dataset or database, the Market Basket Analysis (MBA) approach of DM seeks out a group of items that commonly appear together. This technology is employed in a variety of sectors, including retail, to encourage cross-selling, assist with fraud detection, product replacement, and certain purposes. Based on this technology, it is simple to understand the buying trends of consumers and their preferences. Technology has advanced, and current business practices have significantly changed as a result. By figuring out the connections among the various things in the consumer's buying baskets, this approach examines their purchasing behaviours. Businesses must increase the accuracy of their operations as a result of changes in customer demands. This research focuses on FP growth, which performs better in mining frequent itemsets than apriori. Hence, this paper focuses on analyzing frequent patterns using conditional FP-Tree in FP growth and compares it with improved and traditional apriori with minimum support as the threshold for identifying the frequently occurring item sets. Moreover, the time consumption of the Associative Rule Mining (ARM) model has been compared with the FP Growth algorithm for identifying the short-time comparison model.