Customer-Aware Recommender System for Push Notifications in an e-commerce Environment

Multifaceted nature of personalized services has spawned streams of research. Personalization is done by any ecommerce provider using recommender systems. These systems are coupled with push notification systems to help customers to find the items and collects user behaviour and produce accurate suggestions assisting the system to build a more efficient notification system. This paper proposes a Customer-Aware Recommender System (CUARS) capable of sending push notifications in an ecommerce setting. The developed system is socially responsible and incorporates an analytical approach. CUARS takes advantage of collaborative filtering technique and uses various components like predictive analytics, smart scheduling, feature extraction, ranked factorization recommenders to generate effective push notifications. Personalization plays a key part in maintaining a relevant and unique experience with intuitive recommendations. CUARS can recommend relevant suggestions to the user by assessing their behaviour, profile and notification trends. Recommendations are generated by predictive analysis and refined by profile analysis technique and user trends are incorporated to enhance the notification system. CUARS provides a very efficient way of managing customer segments and reducing churn rate. The performance of the CUARS improves the Click Through Rate by almost 30% and is measured by a comparative analysis with a regular recommendation system assisted by a fixed scheduled and event-based notification system.