Skip to Main Content
Table 3: 
F1 scores for bitext retrieval on BUCC 2017 using random sentences as negative samples instead of nearest neighbors.
MethodDe-EnFr-EnRu-EnZh-En
mBERT w/o ST 47.0 49.3 41.2 38.0 
 
w/ ST (random) 57.7 55.7 48.1 45.2 
w/ ST (hard) 60.6 60.2 49.5 45.7 
MethodDe-EnFr-EnRu-EnZh-En
mBERT w/o ST 47.0 49.3 41.2 38.0 
 
w/ ST (random) 57.7 55.7 48.1 45.2 
w/ ST (hard) 60.6 60.2 49.5 45.7 
Close Modal

or Create an Account

Close Modal
Close Modal