Design and Implementation of a Cross-platform mobile-commerce Recommender System using TensorFlow Recommenders, Express.js, Flutter and Docker Containerization

In mobile commerce, recommendation systems play a pivotal role in bolstering user engagement and fostering business expansion. This study outlines the development of a recommendation system, employing a combination of TensorFlow models, Docker containers, Node.js, and Flutter, tailored specifically for the domain of book recommendations. The challenges inherent in constructing and deploying recommendation systems for both mobile and web applications are systematically addressed in this research. The architecture leverages TensorFlow Recommenders to formulate two essential models: retrieval and ranking. To seamlessly integrate these models into applications, a user-friendly RESTful API is established using Node.js. Additionally, the ongoing development of a prototype mobile application with Flutter demonstrates the practical implementation of this recommendation system. A noteworthy achievement of this paper is the encapsulation of the RESTful API, along with TensorFlow models, within Docker containers as a single deployable image and package. This containerization simplifies the deployment process, enabling the API to be hosted and accessed effortlessly from various platforms, including mobile and web applications, through a designated endpoint. This work amalgamates a diverse range of cutting-edge technologies to fashion a straightforward, yet powerful recommendation system tailored to the unique demands of mobile commerce within the book domain. It underscores the intricacies of recommendation systems and their pivotal role in enriching user experiences. By showcasing the effective utilization of TensorFlow Recommenders, Docker, Node.js, and Flutter, this research offers a practical and robust solution to the complex challenge of constructing, deploying, and seamlessly integrating recommendation systems into the ever-evolving landscape of mobile commerce.