Table 3 summarizes the results of our machine translation evaluation for the Intrinsic-JE and Intrinsic-EJ tasks. In this table, "Proposed" indicates our proposed system, "Baseline" indicates the normal statistical machine translation (moses).
| task | bleu | |||
| Proposed (TORI) | Intrinsic-JE | 0.2565 | ||
| Baseline (moses) | Intrinsic-JE | 0.2156 | ||
| Proposed (TORI) | Intrinsic-EJ | 0.2602 | ||
| Baseline (moses) | Intrinsic-EJ | 0.2501 |
As seen in these results, our method was so effective, as indicated by the BLEU score.