K. Vamshee Krishna, Rishendra Varshith Raju.E, Sai Sri Prabath. Y, Singarweni Rahul 02 7(1) The global videoondemand market is projected to surpass 257 billion by 2027 with platforms like Netflix attributing over 80 of content watched to recommendation engines This highlights the growing demand for intelligent accurate and personalized movie recommendation systems However existing systems face persistent issues such as data sparsity coldstart problems and lack of content diversity which hinder their effectiveness and scalability To address these limitations this research introduces a novel hybrid recommendation framework named ContentBased Collaborative Filtering integrated with Convolutional Neural Networks CBCFCNN The proposed model innovatively combines the strengths of contentbased filtering collaborative filtering and deep learning to deliver a robust and scalable recommendation system CBCFCNN leverages useritem interaction matrices along with highdimensional feature extraction from movie metadata and visual content using CNNs This deep integration allows the system to capture latent patterns and semantic similarities between items thus mitigating the coldstart issue and enriching the recommendation space Furthermore the architecture enhances personalization by learning userspecific behavior from sparse data and enables realtime inference through parallelized CNN processing Unlike traditional models that treat collaborative and contentbased filtering separately CBCFCNN creates a unified representation that dynamically adapts to user preferences while promoting recommendation diversity and relevance The models ability to scale with massive datasets while maintaining low latency and high accuracy makes it suitable for deployment in largescale platforms offering a significant advancement over existing recommendation techniques Copyright (c) Journal of Science Engineering Technology and Management Science. All rights reserved 2026