Deep Learning Structure for Cross-Domain Sentiment Classification Based on Improved Cross Entropy and Weight

Figure 16

Accuracy of each model in Chinese and English datasets. (a) Changes in the accuracy of RNN in the English dataset. (b) Changes in the accuracy of W-RNN in the English dataset. (c) Changes in the accuracy of RNN in the Chinese dataset. (d) Changes in the accuracy of W-RNN in the Chinese dataset.

(a)

(b)

(c)

(d)