A review of recommendation systems
Lobna El Harrouchi, Hanae Moussaoui, Mounir Karmoudi, Nabil El Akkad · 2025
This study aims to highlight the significance of recommendation systems in today's society. It encompasses nearly 95% of our everyday activities, including healthcare, education, social media, e-commerce, employment, and more. The objective of this paper is to explore various methods harnessed in recommendation systems across the domains cited above. A comparison is also made between deep learning, traditional machine learning, and advanced techniques. More than 60 academic articles from journals and conferences are reviewed in this work. In the following parts, we examine how the aforementioned learning techniques and recommendation systems are implemented. Subsequently, we explore potential enhancements to recommendation systems via their integration with complementary AI domains such as NLP, computer vision, and reinforcement learning. Thereafter, we undertake a brief comparative analysis of selected approaches across different datasets. Lastly, we highlight the most widely used datasets for performance assessment in recommendation systems of various methods proposed, analyzing the key datasets and the metrics used to evaluate user preferences.This research facilitates comprehension of recommendation technology, offers information on the approaches that perform well in specific domains or datasets, and directs future research on recommendation system evaluation and optimization using pertinent metrics and datasets.