A Review of Hybrid Recommendation Systems in Online Shopping

Abstract

Hybrid recommendation systems are increasingly used in e-commerce to enhance personalized shopping experiences by overcoming limitations of traditional methods like content-based filtering (CBF) and collaborative filtering (CF). This study reviews 27 scholarly publications to evaluate their effectiveness. Findings indicate that hybrid models improve recommendation accuracy and user satisfaction by integrating multiple filtering techniques. While CBF personalizes recommendations based on past interactions, it struggles with cold start issues. CF, which predicts preferences based on user similarities, faces challenges with sparse data. Hybrid models address these limitations by combining both approaches, dynamically adapting to user interactions, and leveraging multiple data sources. This results in more accurate and engaging recommendations, enhancing user experience and retention.