The Internet's rapid growth has led to information overload, necessitating recommender systems for personalized suggestions. While content-based and collaborative filtering have been successful, data sparsity remains a challenge. To address this, this article presents a novel social recommender system based on convolutional neural networks (SRSCNN). This approach integrates deep learning and contextual information to overcome data sparsity. The SRSCNN model incorporates user and item factors obtained from a neural network architecture, utilizing features from item titles and tags through a CNN. The authors conducted extensive experiments with the MovieLens 10M dataset, demonstrating that the SRSCNN approach outperforms state-of-the-art baselines. This improvement is evident in both rating prediction and ranking accuracy across recommendation lists of varying lengths.