In order to run simple sentiment analysis predictions, use the code below
python run.py -m simple_sentiment -p ./data/CN
Replace "CN" with "EN" and "SG" to do predictions with other dataset.
| Precision | Recall | F1 | |
|---|---|---|---|
| CN | 0.0812 | 0.4929 | 0.1395 |
| EN | 0.5116 | 0.7240 | 0.5996 |
| SG | 0.1950 | 0.5548 | 0.2885 |
| Precision | Recall | F1 | |
|---|---|---|---|
| CN | 0.0393 | 0.2386 | 0.0675 |
| EN | 0.4534 | 0.6416 | 0.5313 |
| SG | 0.1251 | 0.3560 | 0.1851 |
In order to run viterbi sentiment analysis predictions, use the code below
python run.py -m viterbi -p ./data/CN
Replace "CN" with "EN" and "SG" to do predictions with other dataset.
| Precision | Recall | F1 | |
|---|---|---|---|
| CN | 0.2461 | 0.2914 | 0.2668 |
| EN | 0.7446 | 0.8069 | 0.7745 |
| SG | 0.4837 | 0.5155 | 0.4991 |
| Precision | Recall | F1 | |
|---|---|---|---|
| CN | 0.1460 | 0.1729 | 0.1583 |
| EN | 0.6775 | 0.7342 | 0.7047 |
| SG | 0.3945 | 0.4204 | 0.4070 |
In order to run viterbi 3rd best sequence sentiment analysis predictions, use the code below
python run.py -m viterbi_top_k -p ./data/CN
Replace "CN" with "EN" and "SG" to do predictions with other dataset.
| Precision | Recall | F1 | |
|---|---|---|---|
| CN | 0.1368 | 0.3086 | 0.1896 |
| EN | 0.7994 | 0.7564 | 0.7773 |
| SG | 0.4183 | 0.4729 | 0.4440 |
| Precision | Recall | F1 | |
|---|---|---|---|
| CN | 0.0766 | 0.1729 | 0.1062 |
| EN | 0.7638 | 0.7227 | 0.7426 |
| SG | 0.2879 | 0.3255 | 0.3056 |
Run evaluation script by going into eval folder
python evalResult.py <actual data path> <predicted data path>
React
Typescript
Django
Laravel
Facebook
Microsoft
Google
Alibaba
D3
Tencent