Sentiment Classification on SemEval Twitter Sentiment Analysis 2016 (test)
0.648Avg F1 ScoreEnsemble model
Evaluation results
| Method | Links | ||
|---|---|---|---|
| Ensemble modelcomponents=10 CNNs and 10 LSTMs2017.04 | 0.648 | — | |
| CNNEmbedding=word2vec, Convolution size=[5,6,7]2017.04 | 0.646 | — | |
| CNNEmbedding=word2vec, Convolution size=[3,4,5]2017.04 | 0.643 | — | |
| LSTMEmbedding=word2vec, Class weights=false2017.04 | 0.643 | — | |
| CNNEmbedding=word2vec, Convolution size=[1,2,3]2017.04 | 0.642 | — | |
| CNNEmbedding=word2vec, Convolution size=[3,4,5], Fully connected layer=false2017.04 | 0.641 | — | |
| CNNEmbedding=fasttext, Convolution size=[3,4,5]2017.04 | 0.64 | — | |
| CNNEmbedding=word2vec, Convolution size=[3,4,5], Class weights=false2017.04 | 0.64 | — | |
| CNNEmbedding=glove, Convolution size=[3,4,5]2017.04 | 0.637 | — | |
| CNNEmbedding=word2vec, Convolution size=[3,4,5], Distant training=false2017.04 | 0.636 | — | |
| LSTMEmbedding=word2vec2017.04 | 0.636 | — | |
| LSTMEmbedding=word2vec, Fully connected layer=false2017.04 | 0.634 | — | |
| LSTMEmbedding=fasttext2017.04 | 0.633 | — | |
| Previous best historical scores2017.04 | 0.633 | — | |
| LSTMEmbedding=glove2017.04 | 0.63 | — | |
| LSTMEmbedding=word2vec, Distant training=false2017.04 | 0.629 | — | |
| Logistic regressionbaseline=1-3 grams2017.04 | 0.558 | — | |
| BBLSTM-SL2020.04 | — | 0.858 | |
| HAABSA++Method ID=Method 42020.04 | — | 0.87 | |
| LSTM+SynATT+TarRep2020.04 | — | 0.846 | |
| PRET+MULT2020.04 | — | 0.856 | |
| XRCESemEval Winner (SW)=true2020.04 | — | 0.881 |