Skip to Main Content
Table A1 
Conditional Spearman correlation with direct assessment scores for popular and top scoring metrics from the WMT16 Metrics for high-quality and low-quality data partitions. This table corresponds to Table 3 in the main body of the article.
QlowQhighQhigh*All
Meteor 0.297 0.448 0.403 0.565 
-TERp-A 0.255 0.436 0.391 0.554 
MPEDA 0.295 0.444 0.399 0.563 
ROUGE-SU* 0.262 0.407 0.363 0.521 
ChrF3 0.311 0.388 0.339 0.515 
NIST-4 0.242 0.373 0.322 0.478 
BLEU-4 0.173 0.381 0.332 0.456 
-TER 0.133 0.413 0.366 0.455 
-WER 0.095 0.437 0.391 0.443 
-PER 0.177 0.351 0.299 0.429 
 
UPF-Cobalt 0.245 0.437 0.392 0.544 
CP-Oc(*) 0.212 0.391 0.343 0.491 
SP-lNIST 0.258 0.375 0.325 0.483 
DP-Oc(*) 0.108 0.346 0.304 0.385 
SR-Or(*) 0.148 0.196 0.198 0.307 
 
DPMFcomb 0.298 0.471 0.428 0.589 
Metrics-F 0.267 0.477 0.435 0.590 
Cobalt-F-comp 0.234 0.493 0.452 0.586 
BEER 0.302 0.377 0.328 0.505 
UoW-ReVal 0.224 0.423 0.377 0.507 
QlowQhighQhigh*All
Meteor 0.297 0.448 0.403 0.565 
-TERp-A 0.255 0.436 0.391 0.554 
MPEDA 0.295 0.444 0.399 0.563 
ROUGE-SU* 0.262 0.407 0.363 0.521 
ChrF3 0.311 0.388 0.339 0.515 
NIST-4 0.242 0.373 0.322 0.478 
BLEU-4 0.173 0.381 0.332 0.456 
-TER 0.133 0.413 0.366 0.455 
-WER 0.095 0.437 0.391 0.443 
-PER 0.177 0.351 0.299 0.429 
 
UPF-Cobalt 0.245 0.437 0.392 0.544 
CP-Oc(*) 0.212 0.391 0.343 0.491 
SP-lNIST 0.258 0.375 0.325 0.483 
DP-Oc(*) 0.108 0.346 0.304 0.385 
SR-Or(*) 0.148 0.196 0.198 0.307 
 
DPMFcomb 0.298 0.471 0.428 0.589 
Metrics-F 0.267 0.477 0.435 0.590 
Cobalt-F-comp 0.234 0.493 0.452 0.586 
BEER 0.302 0.377 0.328 0.505 
UoW-ReVal 0.224 0.423 0.377 0.507 
Close Modal

or Create an Account

Close Modal
Close Modal