<![CDATA[In an increasingly advanced digital era, the demand for recommendation systems that can provide products that match user preferences is increasingly high. The method used in this research is User-Based Collaborative Filtering, an approach that utilizes the purchasing patterns of other users to provide recommendations to active users. The research process began with collecting transaction datasets and product user preferences from leading e-commerce platforms. This data is then processed through the loading, preprocessing and analysis stages to prepare good data. Model performance is carried out using Mean Absolute Error (MAE) through a 5-fold cross validation process. The evaluation results show that the model has a satisfactory level of accuracy, with an average MAE value of 0.31. This research contributes to the development of a product recommendation system that can help improve users' online shopping experience by providing relevant and personalized recommendations. ]]>
Copyrights © 2024
