Sentiment Score Prediction for New Energy Vehicles Reviews Using Bidirectional LSTM with Bayesian Optimization

The rapid expansion of the new energy vehicles market has increased the need for advanced sentiment analysis tools to extract valuable insights from user reviews. This study introduces a novel sentiment prediction model that combines Bidirectional Long Short-Term Memory networks with Bayesian Optimization to enhance pre-diction accuracy while optimizing computational efficiency. The Bidirectional Long Short-Term Memory model captures both past and future context in sequential data through its bidirectional architecture, while Bayesian Optimization optimizes hyperparameters to improve model performance. Using a dataset of user reviews for new energy vehicles, the performance of the proposed Bayesian Optimization-Bidirectional Long Short-Term Memory model is compared to baseline models such as Long Short-Term Memory, Convolutional Neural Networks, and Gated Recurrent Units. The results show that the model significantly outperforms these baseline models in terms of accuracy and generalization, demonstrating its potential as an effective tool for sentiment score prediction in user feedback.