Enormous data access in social media technologies has become a Himalayan task in the current Global networking field. Many researchers have contributed their contemporary ideas as algorithms in order to balance data accessing and technological executions. People throughout the Globe use technological advancement for the sake of their comfort of living. An ample amount of information is posted on social media. Thus, the users find it difficult to effectively identify the required information. Hence, recommendation systems have become significant factors in the modern e-commerce-based industry. Collaborative Filtering (CF) is one of the important techniques widely used in recommendation systems. However, this technique suffers from the cold-start problem. This paper presents a profile enhancement system that improves the user and item profiles for an effective recommendation. A deep learning-based tree prediction model is used for rating recommendations. Deep Neural Network effectively enhances the prediction process by providing better results. Experimental evaluations were performed with the state-of-the-art of the models. Results obtained and comparisons indicate that the proposed model exhibits low MAE, MSE and RMSE levels, indicating effective predictions.