Research on Text Emotion Analysis Based on LSTM
Xia Liu, Jiaxin Zhou, Ruhua Lu · 2023
Sentiment analysis of text based on deep learning has become an active research direction in the field of natural language processing, where the sentiment analysis of text is used to automatically identify the sentiment tendency embedded in the text, such as positive or negative. Sentiment analysis based on deep learning usually adopts neural network models, which are trained on a large amount of labelled data to capture the sentiment information in the text. The commonly used neural network models include Recurrent Neural Network (RNN), Long Short-Term Memory (LSTM), and so on. In this study, Long Short-Term Memory (LSTM) is used to build a model to construct a text sentiment analyzer, which, after testing 23982 textual data, can predict whether the sentiment expressed is negative or positive according to the input statements, and the accuracy of its prediction results can reach 90.37%. The experimental results show that it is more flexible and generalizable than the traditional rule-based or feature engineering methods, and can further improve the performance of sentiment analysis.