Business Intelligence in Sports Retail: Data Mining Application

The primary research goal is to understand the transactional sales data, in particular, to catch the principal characteristics of the sales, in terms of quantities, sales amount, and invoices similarities, and to find patterns in customer purchases. This project explores the practical applications of Business Intelligence (BI) and Data Mining to sports brand retailers. Starting from an extensive database of customer transactions, where each transaction consists of an item purchased by a customer in a visit, this project applied Business Intelligence (BI) solutions and Data Mining through MySQL and RapidMiner. The analysis outcome is a deeper understanding of interrelated products customer purchase behaviour (most purchase items; items purchased by the customer; items bought together). The massive amount of information available and the various forms of storage has become an issue causing difficulty in transforming and using the data. Therefore, we apply exploratory information management tools and techniques for seasonal data (single month sales). BI Solutions system adds fast data processing, intelligent correlation analysis, a multidimensional analysis target, and the most specific decision-making data through continuous analysis and progressive selection. Cross-selling, collaborative filtering and even up-selling can be easily used and revisit store layout improving store customer service and experience. A deeper understanding of customer behaviour, interrelationships between products as customers purchase them, and recommendations to improve the strategy.