Advancements and Innovations in Recommendation Systems: From Traditional Algorithms to Deep Learning Evolution
Ningxin Tan · Advances in intelligent systems research/Advances in Intelligent Systems Research · 2024
In the digital age, recommender systems have become instrumental in managing information overload by delivering personalized content recommendations.This paper conducts a thorough review of the evolution of deep learning techniques in recommender systems, tracing their development from the initial collaborative filtering methods to the sophisticated use of graph neural networks and knowledge graphs.The study demonstrates that deep learning significantly enhances recommender system capabilities in delivering personalized recommendations, efficiently processing multimodal data, and bolstering user privacy protection.The analysis highlights that early recommender systems primarily relied on collaborative filtering, which, despite its effectiveness, faced challenges such as data sparsity and scalability.The integration of deep learning has revolutionized these systems, enabling the extraction of complex features and patterns from vast datasets.Techniques such as convolutional neural networks (CNNs), recurrent neural networks (RNNs), and transformers have proven effective in capturing nuanced user preferences and item features.Furthermore, the advent of graph neural networks and knowledge graphs has introduced advanced capabilities for handling relational data and incorporating semantic information, significantly improving recommendation accuracy and contextual relevance.