Exploratory Analysis on Market Basket Data using Network Visualization

Abstract

Market basket analysis is a powerful technique for understanding customer behavior and optimizing business strategies based on that understanding. Market-based analysis over time using visualization techniques can provide insights into market trends and relations, simplify complex data, and communicate insights effectively, which can help organizations make more informed decisions. This paper leverages a dataset focused on the users’ incomes and temporal aspects of market purchases. We modeled this dataset as three distinct temporal networks and performed an exploratory evaluation identifying patterns and anomalies in the data. More specifically, we identified groups of related products, indicating thematic purchases, and evaluated the impact of demographic factors, such as income, on customer spending.