M-DA: A Multifeature Text Data-Augmentation Model for Improving Accuracy of Chinese Sentiment Analysis

Table 7

The comparison results of multifeature merging models.

Model	Accuracy	Precision	Recall	F1
(Char_4tag_pos: Word)-BiLSTM-attention	91.53	93.23	89.50	91.34
(Char_4tag_pos dot word)-BiLSTM -attention	91.40	90.82	92.10	91.43
(Char: 4tag_pos: Word)-BiLSTM-attention	92.35	93.10	91.40	92.25
(Char_4tag_pos: Word: Char)-BiLSTM-attention	92.66	93.87	91.20	92.53

(i) The meaning of bold in “model” column: just to make the model name more visually prominent. (ii) The meaning of the bold values provided in the “Accuracy\Precision\Recall\F1” columns: just to highlight the highest value in each column.