Development of a recommendation system for the selection of network equipment based on the Java technology stack

Abstract

The emergence of e-commerce and the growth of the amount of available content creates an overload of large volumes of information. At the same time, recommender systems, being powerful data filtering tools and using a variety of algorithms and analysis methods, reduce this overload by generating the most relevant elements for a particular user, which contributes to more effective and efficient selection decisions. The article discusses the main types and methods of recommender systems, methods for calculating the similarity coefficient of users and elements, metrics for evaluating the quality of work of recommender systems. The main problems of recommender systems are highlighted. The means of development of the client and server part of the system of recommendations for the web application of finding the optimal configuration of network equipment are considered, such as: HTML, CSS, JavaScript, Spring, Spring Boot, Spring Data JPA, Spring Security, Thymeleaf, MySQL. The architecture of a recommender system web application built using these tools is described.