HYBRID PRODUCT RECOMMENDATION SYSTEM USING CONTENT BASED FILTERING AND NEURAL COLLABORATIVE FILTERING
Loading...
Files
Date
Authors
Journal Title
Journal ISSN
Volume Title
Publisher
Universiti Malaysia Sarawak (UNIMAS)
Abstract
Description
With the rapid growth of e-commerce platforms, users are faced with an overwhelming number of product choices. This has created a pressing need for more intelligent and personalized recommendation systems to enhance user experience and drive engagement. Motivated by this trend, this project aims to develop a hybrid recommendation system that combines Content-Based Filtering (CBF) and Neural Collaborative Filtering (NCF) to deliver accurate and relevant product suggestions. The system leverages item attributes such as category and brand in the CBF component, while the NCF model captures deeper user-item interaction patterns. A custom e-commerce prototype was built using Flask to demonstrate how the hybrid model can be integrated into a real-world application. The model was evaluated using metrics like NDCG, precision, and recall, as well as time efficiency to ensure practical usability. Despite limitations such as data quality and sparsity, the project successfully showcases the potential of hybrid approaches in improving recommendation quality. Future improvements may include incorporating more diverse data sources and refining the model for better scalability and performance.