Architecture of a recommendation service for choosing training areas at a higher education institution by applicants with the method of collaborative filtering of machine learning

Abstract

The purpose of the scientific article is to present a study of a recommendation service architecture designed to help applicants in choosing the training area at a higher education institution. The main function of the service is to provide applicants with personalised recommendations for training based on their preferences, interests, academic achievements and ranking of the institution. The architecture is founded on the principle of client-server interaction when clients can receive personalised recommendations and interact with the service through a web interface. The article accomplished the following objectives: the architectural decomposition and description of the main components of the service are completed; a machine learning method is presented, including a collaborative filtering algorithm that is used in the service and allows to consider preferences and offers of other applicants with similar interests and educational profile; recommendations for choosing a user interface for convenient interaction with the service have been developed; test cases have been conducted to assess the effectiveness of the recommendation service. The study shows that the use of the collaborative filtering method in the service architecture makes it possible to achieve high accuracy and satisfaction of applicants when providing recommendations on choosing the training area at a higher education institution. The article has practical significance, as it represents a real application of the machine learning method and service architecture to help applicants choose the field of study. The research results may be useful for the development of similar services in the educational field.