Analyzing Purchase Behavior Using FP Growth Technique to Find Association Rules
Shutchapol Chopvitayakun, Watchareewan Jitsakul, Nuntaporn Aukkanit · 2024
This study examines trends in consumer buying behavior in the context of a supermarket using the FP-GROWTH algorithm, a data mining tool. Transactions including a set of 16 specified items are included in the dataset that is being analyzed: apple, bread, butter, cheese, corn, dill, eggs, ice cream, kidney beans, milk, nutmeg, onion, sugar, unicorn, yogurt, and chocolate. The analysis revealed a notable association rule indicating a 72.1% likelihood of customers purchasing bread when they also buy ice cream and yogurt. This rule is supported by 25.3% of the transactions in the dataset, signifying that a quarter of all purchases involve these three products. Furthermore, the confidence level of this rule is 0.92, suggesting that the probability of customers buying bread in conjunction with ice cream and yogurt is 92% higher compared to those who do not purchase the latter two items. Additional findings include similar association rules, such as a 70.5% likelihood of bread purchases when yogurt and chocolate are bought together, and a 70.6% probability of buying bread alongside sugar and chocolate. These results underscore a significant correlation among the purchases of yogurt, chocolate, sugar, ice cream, and bread, revealing potential strategies for targeted marketing campaigns or product placement optimizations. For instance, retailers could incentivize bread sales through discounts to customers purchasing yogurt, chocolate, sugar, or ice cream, or by strategically positioning bread in close to these items within the store layout to enhance visibility and accessibility.