Sentiment Analysis of Chinese Classic Literary Works Based on Natural Language Processing

This paper discusses the BERT model, two-channel convolutional neural network and bi-directional LSTM model architecture in natural language processing. And on this basis, the BERT analytic representation model was used to extract and analyze the theme of chapter contents and related characters in literary works. Sentiment analysis was conducted with Jia Baoyu in Dream of the Red Chamber as the research object to visualize the sentiment of the plot line of the literary work. Then the classic Chinese literary works exemplified by Dream of Red Mansions, White Deer Plain and Thunderstorm were used as the analytical corpus to analyze the emotional tendencies of the multi-round dialogues between different people in the text, and the emotion vectors were used as inputs of the neural network, which was trained and predicted to get the global emotion matrices between the characters. The experimental results show that the sentence-level emotion judgment accuracy of all three works under the fusion of CNN and bidirectional LSTM algorithms to train the emotion discrimination model is over 85%, while the global emotions of the multiple characters in the text can also be well recognized.