Enhanced Movie Recommendation System Using Neural Collaborative Filtering and Transformer-Based Sentiment Analysis

The present research extreme to construct on later progressions in deep learning to help experts within the plan, advancement, and generation of such a complicated. Suggestion framework for movies. The major commitment of this paper is to highlight the significance of Transformer- based sentiment analysis (BERT/GPT) and Neural Collaborative. Filtering. Unlike traditional collaborative filtering approaches, however, this algorithm provides contextual-aware personalized movie recommendations through blending structured user-item interactions and unstructured user opinions from social media. Using a BERT-CNN sentiment analysis approach, audience sentiments in real-time from microblogging data are used with the NCF framework, which is used for adequate modelling of any interactions between users and objects, in its own aspect. The results obtained from experiments demonstrated the effectiveness of the method. The hybrid model achieves extremely high recall and accuracy in personalized recommendation tasks. Besides the precision- recall curves and sentiment score distribution, the model reaches a decent trade-off between sentiment-based customization and deep recommendation suggestions. The study presents the ability of the proposed hybrid recommendation to develop personalized material dissemination along with its aptitude in establishing an advanced and extremely constructive movie recommendation system.