Sentiment analysis and evaluation of British and American literature based on deep learning

In the context of digitization and information age, deep learning technology has made remarkable progress in the sentiment analysis and evaluation of British and American literature. This study proposes a text sentiment analysis method based on Multilingual BERT with Attention-based Long Short-Term Memory (MLBERT-ALSTM) model, which effectively extracts the dependency of sentiment word context by introducing two self-supervised tasks MBWP and LST. On this basis, this study uses the Attention-based Long Short-Term Memory (ALSTM) layer combined with the attention mechanism for feature extraction to give higher weights to the emotional word region, so as to improve the feature extraction ability of the model. At the same time, in order to retain the relationship characteristics between elements in the input sequence, the article vector is used as the input and the hidden layer vector is generated back to Hierarchical Attention-based Convolutional Neural Network (HACNN). After that, the Text Universal Pooling (TUP) pooling method is used to retain the deep emotional features and improve the accuracy and generalization ability of the model. The experimental results show that the sentiment analysis method based on MLBERT- ALSTM model has significant advantages in extracting abnormal sentiment polarity features, which provides a more effective and accurate tool for the sentiment analysis and evaluation of British and American literature. This research result is helpful to deeply understand and mine the emotional connotation of British and American literary works, and provide a new perspective and method for literary criticism and analysis.