P07-1040 |
Currently , the most widely used
|
automatic MT evaluation
|
metric is the NIST BLEU-4 ( Papineni
|
N04-1022 |
AMTA , 2003 ) . We expect new
|
automatic MT evaluation
|
metrics to emerge frequently
|
P08-1007 |
is invaluable . Among all the
|
automatic MT evaluation
|
metrics , BLEU ( Papineni et
|
N13-3005 |
. Crucially , current standard
|
automatic MT evaluation
|
metrics also lack any diagnostic
|
P05-1067 |
evaluated using the NIST and Bleu
|
automatic MT evaluation
|
software . The evaluation shows
|
P05-1067 |
system using the NIST and Bleu
|
automatic MT evaluation
|
software . The result shows that
|
N07-1005 |
system parameters based on the
|
automatic MT evaluation
|
measures . Acknowledgments The
|
N06-1057 |
translation ( MT ) community for
|
automatic MT evaluation
|
. A problem with ROUGE is that
|
P08-1007 |
et al. , 2007 ) , a total of 11
|
automatic MT evaluation
|
metrics were evaluated for correlation
|
N06-1058 |
correlate with its utility for
|
automatic MT evaluation
|
. Our results suggest that researchers
|
P07-1038 |
an alternative perspective on
|
automatic MT evaluation
|
that may be informative in its
|
N04-1036 |
translation quality is measured by the
|
automatic MT evaluation
|
metrics , such as NIST and Bleu
|
P04-1077 |
framework that automatically evaluated
|
automatic MT evaluation
|
metrics using only manual translations
|
P08-1007 |
judgements than all of the 11
|
automatic MT evaluation
|
metrics evaluated during the
|
P08-1007 |
, having a robust and accurate
|
automatic MT evaluation
|
metric that correlates well with
|
N09-2006 |
riddled error surface computed by
|
automatic MT evaluation
|
metrics . We showed , empirically
|
P08-1007 |
with human judgements than all 11
|
automatic MT evaluation
|
metrics that were evaluated during
|
P06-2003 |
trans - lator . Most approaches to
|
automatic MT evaluation
|
implicitly assume that both criteria
|
P05-1067 |
in section 5 with the NIST/Bleu
|
automatic MT evaluation
|
software and the results are
|
I05-5003 |
translations . Fortunately , the
|
automatic MT evaluation
|
techniques commonly in use do
|