Personalized Recommendation Algorithms in Digital Media: A Review of Engagement Effects, Information Cocoons, and Marketing Implications

Artificial intelligence (AI) has become a significant force, demonstrating rapid and sustained growth on global digital media platforms. AI technologies on these platforms include machine learning, deep learning, natural language processing, expert systems, and highly advanced personalized recommendation algorithms. These AI-driven algorithms are crucial for enhancing user engagement on digital media platforms, such as smartphone applications, social media, and online media. These algorithms provide customized content based on user preferences, significantly influencing user consumption behavior. Meanwhile, the role of recommendation algorithms in content management and creation promotes the boundaries of personalization, which is effective in social media marketing and e-commerce. However, the rapid development of AI-based recommendation algorithms is likely to pose potential risks, such as the "information cocoon effect," producing homogeneous content and exacerbating social polarization and biases. Additionally, concerns about user privacy and ethical implications have been raised. This review aims to explore the multi-faceted impacts of AI-driven personalized recommendation algorithms on audiences of digital media platforms, focusing on audience engagement, changing user behavior through social media marketing, and the impact of the cocoon effect on users.