Navigating Preferences With Artificial Intelligence

Abstract

Artificial Intelligence (AI) has enhanced the ability to deliver personalized suggestions through Recommender Systems (RS), which rely on machine learning, data mining, and deep learning. This study explores the link between AI and RS, examining how AI improves recommendation quality and where challenges remain. Using the MovieLens 10K dataset, we conduct a comparative analysis of widely used algorithms. The performance of RS algorithms is assessed both as a machine learning problem, through collaborative and content-based filtering, and as a deep learning problem, by evaluating recent neural models. We also discuss evaluation methodologies, data types, and the potential of combining algorithms to enhance performance. The findings provide a clearer view of AI's role in RS, highlighting strengths, limitations, and future research opportunities.