OPTIMIZE TEXTILE BOOK RECOMMENDATION SYSTEM USING DEEP LEARNING ALGORITHMS

Abstract

The research aims to optimize the recommendation system for textile books by applying deep learning algorithms. The textile industry, rich in content and material variation, requires a system of recommendations that can accurately accommodate the diverse needs of its users. Deep learning, with its sophistication in processing large and complex data, offers solutions in improving the quality of recommendations. The study explores the use of deep learning models in interpreting user preferences and book characteristics, with the hope of producing more relevant and personal predictions. Research methods that literature conducts systematically through the collection of data from scientific sources such as journals, conferences, and related articles published in the last decade. The results show that deep learning algorithms such as Convolutional Neural Networks (CNN) and Recurrent Neural Network (RNN) have been successfully applied in improving the accuracy of book recommendation systems, including in textile contexts. These models are able to understand and process textile information and user preferences more deeply than traditional algorithms. The research also revealed important factors that influence model performance, such as data quantity and quality, model architecture, and parameter setting. Although there are limitations associated with resource use and the need for large datasets, the use of deep learning algorithms in recommendation systems for textile books shows significant potential in improving personalization and user satisfaction.