“…Guangjun Liu [12] et al proposed a bidirectional GRU-based sentiment analysis algorithm model, which optimizes and improves algorithms by integrating features of sentiment analysis. Yao Lisha [13] et al introduced a deep learning model, the CNN + ATT model, by combining convolutional neural network (CNN) models with attention mechanisms. Mamoona Humayun [14] et al proposed an analysis using class labels to address multi-class problems by combining class labels with similar overall sentiments.…”