Penerapan Data Mining Untuk Prediksi Permintaan Hasil Pertanian Beras Menggunakan Metode Fp-Growth Berbasis Website

Abstract

This research aims to implement the FP-Growth algorithm in predicting rice demand through a web-based system at Toko Pangan Kita in Pesisir Selatan Regency. The FP-Growth method is applied to identify frequent purchasing patterns from transaction data, thereby supporting business owners and warehouse managers in making strategic decisions related to stock management and distribution. The system was developed using PHP and MySQL and equipped with features such as registration, login, and password reset to ensure user management and security. The analysis process was conducted through two approaches, namely the manual implementation of FP-Growth via programming and validation using RapidMiner. The results show that the FP-Growth algorithm can generate association rules with varying support and confidence values, such as the rule Ir 42 > Solok with 36% support and 79% confidence, which indicates a very strong relationship. In conclusion, this prediction system proves effective in analyzing rice sales data at Toko Pangan Kita and provides strategic insights for business management, thus enhancing stock management efficiency, improving customer service, and supporting more accurate rice distribution planning.