D15-1248 |
substantial improvements from
|
compound splitting
|
of 0.7 -- 1.1 BLEU . On newstest2014
|
E09-3008 |
et al. ( 2006 ) suggested using
|
compound splitting
|
to improve alignment , or to
|
E06-1006 |
Finnish . A variety of ways for
|
compound splitting
|
have been investigated in machine
|
D11-1089 |
The largest obstacle that makes
|
compound splitting
|
difficult is the existence of
|
E03-1076 |
into English as action plan .
|
Compound splitting
|
is a well defined computational
|
D11-1089 |
augmenting discriminative models of
|
compound splitting
|
with large external linguistic
|
C00-2162 |
prefixes are treated in addition to
|
compound splitting
|
. Experiments for POS-annotation
|
D11-1089 |
proposed the use of query logs for
|
compound splitting
|
.3 Their experimental results
|
E03-1076 |
data-driven method that combines
|
compound splitting
|
and word recombination for speech
|
E03-1076 |
score . The words resulting from
|
compound splitting
|
could also be marked as such
|
D15-1248 |
particle verbs . Bi - narization ,
|
compound splitting
|
, and particle verb restructuring
|
D11-1089 |
baseline , UNI - GRAM , performs
|
compound splitting
|
based on a word 1-gram language
|
E03-1076 |
One way to define the goal of
|
compound splitting
|
is to break up foreign words
|
D11-1089 |
and word segmentation ( or noun
|
compound splitting
|
) was not at all discussed .
|
E12-1068 |
produce a gain in performance . For
|
compound splitting
|
, we follow Fritzinger and Fraser
|
D11-1089 |
such compounds . For example ,
|
compound splitting
|
enables SMT systems to translate
|
C00-2162 |
training reduces from 150 to 81 by
|
compound splitting
|
and can further be reduced to
|
E06-1006 |
operations , in particular stemming and
|
compound splitting
|
, are interleaved such that a
|
E03-1076 |
<title> Empirical Methods for
|
Compound Splitting
|
</title> Philipp Koehn Kevin
|
E03-1076 |
use lexicon based approaches to
|
compound splitting
|
for information retrieval . Compounds
|