Pomade Product Recommendation System Usingthe Collaborative Filtering Method

Abstract

The rapid growth of digital technology has driven the adoption of recommendation systems across various platforms to assist users in finding products that match their preferences. This study develops a pomade product suggestion system by applying the Collaborative Filtering method using a User-Based approach. The choice of pomade as the study object is based on the increasing interest in men's hair care products and the wide variety of options available on e-commerce platforms, which often confuse consumers. The research data were collected through a web scraping process from Tokopedia, resulting in 4,375 user ratings for 20 pomade products. The preprocessing stage included data transformation, duplicate removal, and label encoding. User similarity was calculated using the cosine similarity method to identify groups of users with similar preference patterns. The system was evaluated using the Mean Absolute Error (MAE) metric to measure prediction accuracy. The results indicate that the system can provide relevant recommendations with an acceptable level of accuracy. These findings suggest that implementing Collaborative Filtering for pomade products can be an effective solution to assist consumers in selecting suitable products, while contributing to the advancement of recommendation technology in e-commerce.