Cross-lingual Document Classification on Reuters de -> en
76.9AccuracyBICVM+
Evaluation results
| Method | Links | |
|---|---|---|
| BICVM+Number of labeled examples=1000, Classifier=Averaged perceptron2013.12 | 76.9 | |
| BilBOWATraining Time (min)=6, Embedding Dimension=402014.10 | 75 | |
| Bilingual Auto-encoders (BAEs)Training Time (min)=4,800, Embedding Dimension=402014.10 | 72.8 | |
| BICVMNumber of labeled examples=1000, Classifier=Averaged perceptron2013.12 | 71.4 | |
| BiCVMTraining Time (min)=15, Embedding Dimension=402014.10 | 71.4 | |
| I-MatrixNumber of labeled examples=1000, Classifier=Averaged perceptron2013.12 | 71.1 | |
| Klementiev et al.Training Time (min)=14,400, Embedding Dimension=402014.10 | 71.1 | |
| GlossedNumber of labeled examples=1000, Classifier=Averaged perceptron2013.12 | 68.6 | |
| Glossed BaselineEmbedding Dimension=402014.10 | 68.6 | |
| MTNumber of labeled examples=1000, Classifier=Averaged perceptron2013.12 | 67.4 | |
| MT BaselineEmbedding Dimension=402014.10 | 67.4 | |
| Majority ClassNumber of labeled examples=1000, Classifier=Averaged perceptron2013.12 | 46.8 | |
| Majority BaselineEmbedding Dimension=402014.10 | 46.8 |