VR Supermarket: a Virtual Reality Online Shopping Platform with a Dynamic Recommendation System

In this paper, VR Supermarket, a Virtual Reality (VR) online shopping platform is proposed to provide an immersive shopping experience for users. The system consists of four components – the VR application, front-end interfaces, the external databases and the recommendation system. The front-end interfaces handle the user input in a 360° rendered showroom where users can explore and buy products. Structured Query Languages (SQL) and Not only SQL (NoSQL) databases are integrated with the application to store inventory, customer information and billing information. An integrated recommendation system provides users a personalized shopping experience by recommending relevant products. The recommendations are based on the purchase history, which makes the VR supermarket a dynamic, adaptive and user-oriented system. The paper proposes a Neural Collaborative Filtering (NCF) recommendation system augmented by Convolutional Neural Networks (CNN) and Attention layers. The VR supermarket provides a realistic user-friendly shopping experience where users are able to select from a variety of products and place orders. Seamless integration with databases enables extensive storage of supermarket data and the potential to scale up the application into an enterprise-level application that will add a new dimension to the statement “Shop at home”.