Table 3 Comparison of character-level BLI methods from prior work [44, 45] with automatically learned character-level representations
From: A deep learning approach to bilingual lexicon induction in the biomedical domain
| Â | Development | |||||
|---|---|---|---|---|---|---|
| Â | Words | Phrases | Words + Phrases | |||
Representation | F1 (top) | F1 (all) | F1 (top) | F1 (all) | F1 (top) | F1 (all) |
ED norm | 24.49 | 19.53 | 15.62 | 19.87 | 23.83 | 19.55 |
log(ED rank) | 28.57 | 28.17 | 18.05 | 17.27 | 27.86 | 27.46 |
ed norm+ log(ED rank) | 25.99 | 11.20 | 18.40 | 14.35 | 25.49 | 11.31 |
CHARPAIRS | 31.95 | 32.32 | 23.70 | 25.97 | 31.39 | 31.92 |
| Â | Test | |||||
| Â | Words | Phrases | Words + Phrases | |||
Representation | F1 (top) | F1 (all) | F1 (top) | F1 (all) | F1 (top) | F1 (all) |
ED norm | 28.10 | 28.29 | 8.70 | 8.63 | 26.97 | 27.24 |
log(ED rank) | 29.30 | 28.95 | 19.48 | 19.35 | 28.70 | 28.39 |
ed norm+ log(ED rank) | 29.76 | 29.65 | 17.57 | 17.45 | 29.05 | 29.00 |
CHARPAIRS | 30.70 | 32.19 | 31.82 | 30.61 | 30.81 | 32.15 |