This project compare different advances techniques for Sentiment Analysis in Spanish Tweets using data samples of Sepln-TASS
| Model | Macro F1 | Accuracy |
|---|---|---|
| CNN | 0.330 | 0.77075 |
| SIF | 0.733 | 0.773 |
| ULMFit | 0.222 | 0.3837 |
| Model | Run | Macro F1 | Accuracy |
|---|---|---|---|
| CNN | cnn-btf-2019_03_13-04_12_55 | 0.387 | 0.432 |
| SIF | SIF-M LP-2019_01_30-00_27_52 | 0.399 | 0.398 |
| ULMFit | md3-2019_03_12-06_37_15 | 0.270 | 0.299 |
| Model | Accuracy |
|---|---|
| CNN | 0.755 |
| SIF | 0.602 |
| ULMFit | 0.309 |
Links to the embeddings (#dimensions=300, #vectors=1000653)
- Implementation: Word2Vec with Skipgram by GenSim
- Parameters: For details on parameters please refer to the SBWCE page
- Spanish Billion Word Corpus
- Corpus Size: 1.4 billion words
Word embeddings were computed by Cristian Cardellino. Please refer to the SBWCE page if you want to use these vectors.