<![CDATA[In today's digital era, recommendation systems have become an integral part of supporting consumer purchasing decisions, including in the food and beverage industry. This study aims to develop a product recommendation system for snacks and beverages using the item-based collaborative filtering method. This method was chosen due to its ability to handle large-scale user and product data, as well as its efficiency in providing relevant recommendations based on user consumption patterns. In this study, the system calculates the average user rating and implements Cosine Similarity to measure the similarity between products, resulting in more accurate recommendations. The system also evaluates the accuracy of recommendations using the Mean Absolute Error (MAE) metric. Based on the results obtained, which is 0.285403 for the average error on 17 items, the developed recommendation system can improve consumers' shopping experience, help them find products that suit their tastes, and support the sales of snacks and beverages products in the market]]>
Copyrights © 2025
