Abstract :

The rapid growth of online news platforms has resulted in an overwhelming amount of content, making it challenging for users to access relevant news. Traditional text-based recommendation systems often fail to capture the visual context of news, which can play a critical role in user engagement. This paper proposes an AI-driven news recommendation system that leverages image analysis along with textual content to improve personalization and accuracy. By extracting features from news images using convolutional neural networks (CNNs) and combining them with natural language processing (NLP) of headlines and descriptions, the system recommends relevant news items to users. Experimental results show enhanced user engagement and higher recommendation accuracy compared to traditional approaches, demonstrating the efficacy of integrating visual information into news recommendation systems.