Traditional Movie recommendation systems (MRS) typically depends on either content-based filtering or collaborative filtering methods, which often results in generic suggestions lacking in person-specific and accuracy. Our research work addresses the limitations of convention-al movie recommendation systems. The proposed solution integrates both ‘Content-based and Collaborative Filtering’ methodologies to enhance the recommendation accuracy based on individual preferences. Techniques such as Cosine Similarity, K-Nearest Neighbour (KNN), K-means Clustering, and small-scale natural language processing (NLP) are employed to im-prove diversity and precision of recommendations. The use of a hybrid approach which uses Item and User-based collaborative filtering methods for improving recommendations through similarities and preferences of movies. It describes how the use of machine learning algorithms can be incorporated in the process of training and updating collaborative dataset, enabled by the Data Refresh Mechanism, for achieving reliability of movie suggestions by handling data con-flicts. In terms of Content Based filtering, the TMDB 5000 Movies Dataset will be used instead of Collaborative Filtering. Comparisons of different models have been made, which reveals how Content-based and Collaborative-based Filtering perform effectively in terms of recommenda-tions accuracy around 87.5%. It should be noted that while comprehensive results are presented for Content Based Filtering, limitations in resources and the complexity of implementing and combining both approaches may restrict the depth of results for Collaborative Filtering. Addi-tionally, insights into the performance of the hybrid model and strategies for managing data loss during refresh cycles are discussed. The future applications of our proposed model are Enhanced Personalization in Streaming Services, Improvement in E-commerce and Marketing, Applicabil-ity in social media and Content Platforms and Academic and Research Implications.