Unveiling movie emotions: BERT-BiLSTM fusion and VADER models for enhanced sentiment analysis
C. Sai Manasa, Salim Salim, A. Shreyank, Ankur Vaibhav · 2025
In the domain of sentiment analysis, understanding the nuanced emotions and opinions embedded within movie reviews holds significant importance for various applications such as recommendation systems, marketing strategies, and audience engagement. This paper presents a novel approach to enhance sentiment analysis in the context of movie reviews by leveraging the synergies between Bidirectional Encoder Representations from Transformers (BERT) with Bidirectional Long Short-Term Memory (BiLSTM) networks and the valence-aware dictionary and sentiment Reasoner (VADER) model. Through a comprehensive evaluation on benchmark movie review datasets, we demonstrate that the fusion of BERT-BiLSTM with VADER results in improved sentiment analysis performance compared to existing methods and illuminates the intricate emotional landscapes depicted in movie reviews.