Mikel Artetxe, Holger Schwenk
We introduce an architecture to learn joint multilingual sentence representations for 93 languages, belonging to more than 30 different families and written in 28 different scripts. Our system uses a single BiLSTM encoder with a shared BPE vocabulary for all languages, which is coupled with an auxiliary decoder and trained on publicly available parallel corpora. This enables us to learn a classifier on top of the resulting embeddings using English annotated data only, and transfer it to any of the 93 languages without any modification. Our experiments in cross-lingual natural language inference (XNLI dataset), cross-lingual document classification (MLDoc dataset) and parallel corpus mining (BUCC dataset) show the effectiveness of our approach. We also introduce a new test set of aligned sentences in 112 languages, and show that our sentence embeddings obtain strong results in multilingual similarity search even for low-resource languages. Our implementation, the pre-trained encoder and the multilingual test set are available at https://github.com/facebookresearch/LASER
| Task | Dataset | Metric | Value | Model |
|---|---|---|---|---|
| Cross-Lingual | MLDoc Zero-Shot English-to-Japanese | Accuracy | 60.3 | Massively Multilingual Sentence Embeddings |
| Cross-Lingual | MLDoc Zero-Shot English-to-French | Accuracy | 77.95 | Massively Multilingual Sentence Embeddings |
| Cross-Lingual | MLDoc Zero-Shot English-to-Chinese | Accuracy | 71.93 | Massively Multilingual Sentence Embeddings |
| Cross-Lingual | MLDoc Zero-Shot English-to-Spanish | Accuracy | 77.33 | Massively Multilingual Sentence Embeddings |
| Cross-Lingual | MLDoc Zero-Shot English-to-Russian | Accuracy | 67.78 | Massively Multilingual Sentence Embeddings |
| Cross-Lingual | MLDoc Zero-Shot English-to-Italian | Accuracy | 69.43 | Massively Multilingual Sentence Embeddings |
| Cross-Lingual Bitext Mining | BUCC French-to-English | F1 score | 93.91 | Massively Multilingual Sentence Embeddings |
| Cross-Lingual Bitext Mining | BUCC German-to-English | F1 score | 96.19 | Massively Multilingual Sentence Embeddings |
| Cross-Lingual Bitext Mining | BUCC Chinese-to-English | F1 score | 92.27 | Massively Multilingual Sentence Embeddings |
| Cross-Lingual Bitext Mining | BUCC Russian-to-English | F1 score | 93.3 | Massively Multilingual Sentence Embeddings |
| Cross-Lingual Document Classification | MLDoc Zero-Shot English-to-Japanese | Accuracy | 60.3 | Massively Multilingual Sentence Embeddings |
| Cross-Lingual Document Classification | MLDoc Zero-Shot English-to-French | Accuracy | 77.95 | Massively Multilingual Sentence Embeddings |
| Cross-Lingual Document Classification | MLDoc Zero-Shot English-to-Chinese | Accuracy | 71.93 | Massively Multilingual Sentence Embeddings |
| Cross-Lingual Document Classification | MLDoc Zero-Shot English-to-Spanish | Accuracy | 77.33 | Massively Multilingual Sentence Embeddings |
| Cross-Lingual Document Classification | MLDoc Zero-Shot English-to-Russian | Accuracy | 67.78 | Massively Multilingual Sentence Embeddings |
| Cross-Lingual Document Classification | MLDoc Zero-Shot English-to-Italian | Accuracy | 69.43 | Massively Multilingual Sentence Embeddings |