Enhancing Movie Recommendation Systems with BERT: A Deep Learning Approach

The proposed paper introduces an innovative approach to enhance movie recommendation systems by integrating BERT. The approach employs BERT's advanced text processing to improve the understanding of movie description. Through experiments comparing traditional algorithms with BERT-augmented ones, significant improvements in recommendation precision are demonstrated. The proposed model combines BERT to identify similar movies and TF-IDF for additional suggestions, creating an efficient recommendation system. A comparative study of Word2Vec, BoW, and BERT strengthens the research foundation, highlighting BERT's superior performance by 11% and 8%, respectively.