Table 8.
Sentiment classification on different datasets based on different situations (AFINN and NO_AFINN).
| Android | |||
|---|---|---|---|
| Metric | Metric\Dic | AFINN | NO_AFINN |
| WJST | Accuracy1 | 0.7425 | 0.5725 |
| Accuracy2 | 0.7375 | 0.58 | |
| Perplexity | 15.53 | 16.1551 | |
| Topic_Coh | −2.2654 | −1.7346 | |
| WJST1 | Accuracy1 | 0.855 | 0.81 |
| Accuracy2 | 0.8475 | 0.8075 | |
| Perplexity | 16.3399 | 16.0482 | |
| Topic_Coh | −2.2228 | −0.1295 | |
|
| |||
| Automotive | |||
| WJST | Accuracy1 | 0.7125 | 0.6025 |
| Accuracy2 | 0.7025 | 0.6075 | |
| Perplexity | 20.4488 | 20.4065 | |
| Topic_Coh | −3.2282 | −1.6628 | |
| WJST1 | Accuracy1 | 0.7925 | 0.7025 |
| Accuracy2 | 0.79 | 0.7125 | |
| Perplexity | 20.5213 | 21.1296 | |
| Topic_Coh | −0.326 | −1.1809 | |
|
| |||
| Electronic | |||
| WJST | Accuracy1 | 0.8525 | 0.705 |
| Accuracy2 | 0.8425 | 0.6825 | |
| Perplexity | 20.0579 | 20.2615 | |
| Topic_Coh | −0.5322 | −0.5926 | |
| WJST1 | Accuracy1 | 0.8475 | 0.76 |
| Accuracy2 | 0.855 | 0.765 | |
| Perplexity | 21.8195 | 21.5739 | |
| Topic_Coh | −1.5586 | −1.6968 | |
|
| |||
| Movie | |||
| WJST | Accuracy1 | 0.8475 | 0.71 |
| Accuracy2 | 0.8325 | 0.715 | |
| Perplexity | 22.4588 | 22.6124 | |
| Topic_Coh | −0.9342 | −0.4637 | |
| WJST1 | Accuracy1 | 0.9575 | 0.485 |
| Accuracy2 | 0.945 | 0.4875 | |
| Perplexity | 23.5662 | 22.8134 | |
| Topic_Coh | −1.2359 | −1.3717 | |