Sentiment Analysis of E-Commerce Product Reviews Based on BiLSTM Model
Xiqin Ao, Liyang Zhao, Xingyu Liu, Tingting Wu · 2024
In this study, the BiLSTM model is used to improve the accuracy of sentiment analysis. Firstly, CNN, LSTM and their bidirectional version BiLSTM are introduced, highlighting their advantages in processing text time series. The experimental results of constructing a BiLSTM model and conducting verification show that the model is superior to other models in identifying emotional polarity, and has good generalization and high accuracy. This study can not only provide more accurate analysis for the sentiment analysis of e-commerce reviews, but also show the application of deep learning in natural language processing.