Table 6: 

Term usage percentage (Term%) and BLEU scores of En-De models on terminology test sets (Dinu et al., 2019) provided with correct terminology entries (exact matches on both source and target sides). EDITOR with soft constraints achieves higher BLEU than LevT with soft constraints, and on par or higher BLEU than LevT with hard constraints.

WiktionaryIATE
Term%↑BLEU↑Term%↑BLEU↑
Prior Results 
Base Trans. 76.9 26.0 76.3 25.8 
Post18 99.5 25.8 82.0 25.3 
Dinu19 93.4 26.3 94.5 26.0 
Base LevT 81.1 30.2 80.3 29.0 
Susanto20 100.0 31.2 100.0 30.1 
 
Our Results 
LevT 84.3 28.2 83.9 27.9 
+ soft constraints 90.5 28.5 92.5 28.3 
+ hard constraints 100.0 28.8 100.0 28.9 
EDITOR 83.5 28.8 83.0 27.9 
+ soft constraints 96.8 29.3 97.1 28.8 
+ hard constraints 99.8 29.3 100.0 28.9 
WiktionaryIATE
Term%↑BLEU↑Term%↑BLEU↑
Prior Results 
Base Trans. 76.9 26.0 76.3 25.8 
Post18 99.5 25.8 82.0 25.3 
Dinu19 93.4 26.3 94.5 26.0 
Base LevT 81.1 30.2 80.3 29.0 
Susanto20 100.0 31.2 100.0 30.1 
 
Our Results 
LevT 84.3 28.2 83.9 27.9 
+ soft constraints 90.5 28.5 92.5 28.3 
+ hard constraints 100.0 28.8 100.0 28.9 
EDITOR 83.5 28.8 83.0 27.9 
+ soft constraints 96.8 29.3 97.1 28.8 
+ hard constraints 99.8 29.3 100.0 28.9 
Close Modal

or Create an Account

Close Modal
Close Modal