Програмний модуль рекомендаційної системи товарів

Abstract

This work is devoted to the development of a software module of a product recommendation system for online shopping using a hybrid approach to product recommendations. The work implements a clientserver architecture with support for user accounts, collection of browsing history, as well as an algorithm for generating personalized recommendations based on TF-IDF analysis of product descriptions and user behavioral patterns. The software is developed using the Python language, the Flask framework, the Pandas and Scikit-learn libraries. To verify the quality of the developed module, testing was conducted, which includes checking the system components, as well as evaluating the accuracy of recommendations using metrics such as Precision, Recall and F1.