Fig. 3

The confusion matrices (8 × 8) of four different model. A The confusion matrix of TEXT-CNN. B The confusion matrix of LSTM. C The confusion matrix of BERT. D The confusion matrix of HyM. The row indicates the number of data instances belonging to this class, and column of the confusion matrix indicates the number of data instances that has been predicted as this category. The confusion matrix is used to summarize the results of a classifier, where the closer the color is to the dark green in the middle of the picture, the more accurate the model is. From the four aboved pictures, the confusion matrix show that the HyM is superior to traditional classification models (TEXT-CNN, LSTM, BERT). TEXT-CNN, text convolutional neural networks; LSTM, long short-term memory; BERT, bidirectional encoder representations from transformers; HyM, Hybrid Model