Table 1: 
F1 scores for unsupervised bitext retrieval on BUCC 2017. Results with mBERT are from our method (Sec. 2) using the final (12th) layer. We also include results for the 8th layer (e.g., Libovický et al., 2019), but do not consider this part of the unsupervised setting as we would not have known a priori which layer was best to use.
MethodDe-EnFr-EnRu-EnZh-En
Hangya and Fraser (2019) 
avg. 30.96 44.81 19.80 − 
align-static 42.81 42.21 24.53 − 
align-dyn. 43.35 43.44 24.97 − 
 
Our method 
mBERT (final layer) 42.1 45.8 36.9 35.8 
+ digit filtering (DF) 47.0 49.3 41.2 38.0 
+ edit distance (ED) 47.0 49.3 41.2 38.0 
+ self-training (ST) 60.6 60.2 49.5 45.7 
 
mBERT (layer 8) 67.0 65.3 59.3 53.3 
+ DF, ED, ST 74.9 73.0 69.9 60.1 
MethodDe-EnFr-EnRu-EnZh-En
Hangya and Fraser (2019) 
avg. 30.96 44.81 19.80 − 
align-static 42.81 42.21 24.53 − 
align-dyn. 43.35 43.44 24.97 − 
 
Our method 
mBERT (final layer) 42.1 45.8 36.9 35.8 
+ digit filtering (DF) 47.0 49.3 41.2 38.0 
+ edit distance (ED) 47.0 49.3 41.2 38.0 
+ self-training (ST) 60.6 60.2 49.5 45.7 
 
mBERT (layer 8) 67.0 65.3 59.3 53.3 
+ DF, ED, ST 74.9 73.0 69.9 60.1 
View Large
Close Modal
This Feature Is Available To Subscribers Only

or Create an Account

Close Modal
Close Modal
This site uses cookies. By continuing to use our website, you are agreeing to our privacy policy. No content on this site may be used to train artificial intelligence systems without permission in writing from the MIT Press.
Accept