Product recommendation system on building materials shopping using FP-Growth algorithm

Abstract

The application program created is an e-commerce website based application program that is used to provide product choices for building materials using the FP-Growth algorithm. This algorithm is used to find frequent itemset (data that most often appears) on store transaction data from January 2019 to the most recent transaction. The use of FP-Tree which is used simultaneously with the FP-Growth algorithm only scans the database 2 times. The first scanning is done to read the previous store sales transaction data. The second scanning on the tree trace to form a pattern. After obtaining a number of linkages that are higher than the minimum support from the transaction data, product choices can be generated on the website. This application was created using the Hypertext Processor (PHP) programming language base which is supported by the MySQL database and SQL Server. It is expected that the presence of web-based applications can improve store sales and the marketing process of building products for the better.