Chinese Text Sentiment Analysis Model Based on BERT and BiTCN

As one of the research hotspots in natural language processing, text sentiment analysis has received extensive attention in recent years. Aiming at the problem that the unidirectional TCN in BERT-TCN cannot fully extract the contextual semantic features of the word vectors of text, a text sentiment analysis model based on BERT and Bidirectional Temporal Convolutional Network (BiTCN) is proposed. The model first uses the BERT model to obtain the word vector representations of text, and then the word vectors are input into BiTCN to extract contextual semantic features from the forward and backward directions of the word vectors. Finally, the contextual semantic features are input into a softmax classifier for sentiment classification. Experiments show that compared with the BERT-TCN model, the BERT-BiTCN model has improved accuracy, precision, recall, and F1 value.