Conference material: "Proceedings of the International Conference on Computer Graphics and Vision “Graphicon” (19-21 September 2023, Moscow)"
Authors:Pronin D.S., Zarubin I.B.
Informational Recommendatory System of Selection and Ranking of Works of Graphic Digital Art using Combined Collaborative and Content-based Approach
The use of several methods of recommendations in the development of a recommendatory system of works of digital art makes it possible to increase the accuracy and uniqueness of the selected objects. The «Weighted Hybrid Recommendation» methodology combines content-oriented and collaborative methods of constructing recommendations. The content-oriented method analyses features and user profiles, and the collaborative method takes into account user evaluations and feedback. The presented mathematical model allows to dynamically prioritize intermediate predicted estimates depending on the quality and quantity of available data algorithm. The recommendation system is implemented in the Java programming language in the Spring Boot web application hosted on the Linux cloud server. The developed model was tested on the focus group of students. The elements of the web application’s user interface were demonstrated, such as an initial survey, a slider with recommendations, displaying features of objects and the ability to evaluate an object. The technical solutions used, their advantages and disadvantages, as well as alternative technical solutions are described. A further way of development of recommendatory system is outlined.