APPLICATION OF NEURAL NETWORKS IN PERSONALIZED RECOMMENDATION SYSTEMS

This paper explores modern approaches to building personalized recommendation systems using artificial intelligence methods, particularly neural network models. Traditional recommendation methods such as collaborative filtering and content-based models, along with their limitations, are analyzed. Special attention is paid to various types of neural networks, including fully connected networks (FCNN), convolutional networks (CNN), recurrent networks (RNN, LSTM, GRU), and transformers (BERT, GPT). Examples of practical applications of these methods in different domains, such as event recommendations, movies, music, products, and educational content, are provided. The advantages and disadvantages of neural network-based approaches, as well as their future development prospects, are discussed.