Desarrollo de un sistema de recomendación de productos tecnológicos basado en aprendizaje automático para mejorar la experiencia del consumidor

Abstract

Nowadays, the proliferation of technological devices has completely transformed the way consumers access information and make purchasing decisions. The vast variety of products available in the market has created a complex and sometimes misleading environment for users. Added to this is the diversity of needs and preferences that each consumer has on a personal level, making the correct choice of a product a challenging task. In response to this situation, this document analyzes the development and implementation of a product recommendation system as part of the internship for UrU Tecnología. This system is based on advanced machine learning methodologies and robust data acquisition through scraping, followed by ETL (Extract, Transform, Load) processes to enable efficient structuring and optimization of the collected information. By comparing different recommendation models, such as collaborative filtering and machine learning approaches, this study seeks to evaluate the system's ability to generate personalized recommendations tailored to individual user needs. The document also outlines the data preprocessing and model training steps and explains how the developed model can be hosted on Amazon Web Services to achieve a scalable and flexible architecture. The results of this implementation indicated that the UrU recommendation system not only facilitates purchasing decisions but also enhances user satisfaction by providing more relevant recommendations that align with their expectations. Therefore, this work contributes not only to the academic field of data analysis and recommendation systems but also offers an innovative perspective on how technology can simplify decision-making in an increasingly complex commercial environment.