#2544We propose a newphrase-based translation model and decoding algorithm that enables us to evaluate and compare several, previously proposed phrase-based translation models.
tech,8-1-N03-1017,ak
phrase-based translation model
</term>
and
<term>
decoding algorithm
</term>
that enables us to evaluate and compare
#2548We propose a new phrase-based translation model anddecoding algorithm that enables us to evaluate and compare several, previously proposed phrase-based translation models.
model,21-1-N03-1017,ak
compare several , previously proposed
<term>
phrase-based translation models
</term>
. Within our framework , we carry
#2561We propose a new phrase-based translation model and decoding algorithm that enables us to evaluate and compare several, previously proposedphrase-based translation models.
model,18-2-N03-1017,ak
understand better and explain why
<term>
phrase-based models
</term>
outperform
<term>
word-based models
#2583Within our framework, we carry out a large number of experiments to understand better and explain whyphrase-based models outperform word-based models.
model,21-2-N03-1017,ak
phrase-based models
</term>
outperform
<term>
word-based models
</term>
. Our empirical results , which hold
#2586Within our framework, we carry out a large number of experiments to understand better and explain why phrase-based models outperformword-based models.
other,9-3-N03-1017,ak
results , which hold for all examined
<term>
language pairs
</term>
, suggest that the highest levels
#2598Our empirical results, which hold for all examinedlanguage pairs, suggest that the highest levels of performance can be obtained through relatively simple means: heuristic learning of phrase translations from word-based alignments and lexical weighting of phrase translations.
tech,27-3-N03-1017,ak
through relatively simple means :
<term>
heuristic learning
</term>
of
<term>
phrase translations
</term>
#2616Our empirical results, which hold for all examined language pairs, suggest that the highest levels of performance can be obtained through relatively simple means:heuristic learning of phrase translations from word-based alignments and lexical weighting of phrase translations.
other,30-3-N03-1017,ak
:
<term>
heuristic learning
</term>
of
<term>
phrase translations
</term>
from
<term>
word-based alignments
</term>
#2619Our empirical results, which hold for all examined language pairs, suggest that the highest levels of performance can be obtained through relatively simple means: heuristic learning ofphrase translations from word-based alignments and lexical weighting of phrase translations.
lr,33-3-N03-1017,ak
<term>
phrase translations
</term>
from
<term>
word-based alignments
</term>
and
<term>
lexical weighting
</term>
#2622Our empirical results, which hold for all examined language pairs, suggest that the highest levels of performance can be obtained through relatively simple means: heuristic learning of phrase translations fromword-based alignments and lexical weighting of phrase translations.
tech,36-3-N03-1017,ak
<term>
word-based alignments
</term>
and
<term>
lexical weighting
</term>
of
<term>
phrase translations
</term>
#2625Our empirical results, which hold for all examined language pairs, suggest that the highest levels of performance can be obtained through relatively simple means: heuristic learning of phrase translations from word-based alignments andlexical weighting of phrase translations.
other,39-3-N03-1017,ak
and
<term>
lexical weighting
</term>
of
<term>
phrase translations
</term>
. Surprisingly , learning
<term>
phrases
#2628Our empirical results, which hold for all examined language pairs, suggest that the highest levels of performance can be obtained through relatively simple means: heuristic learning of phrase translations from word-based alignments and lexical weighting ofphrase translations.
#2634Surprisingly, learningphrases longer than three words and learning phrases from high-accuracy word-level alignment models does not have a strong impact on performance.
other,7-4-N03-1017,ak
<term>
phrases
</term>
longer than three
<term>
words
</term>
and learning
<term>
phrases
</term>
from
#2638Surprisingly, learning phrases longer than threewords and learning phrases from high-accuracy word-level alignment models does not have a strong impact on performance.
other,10-4-N03-1017,ak
three
<term>
words
</term>
and learning
<term>
phrases
</term>
from
<term>
high-accuracy word-level
#2641Surprisingly, learning phrases longer than three words and learningphrases from high-accuracy word-level alignment models does not have a strong impact on performance.
model,12-4-N03-1017,ak
and learning
<term>
phrases
</term>
from
<term>
high-accuracy word-level alignment models
</term>
does not have a strong impact on
#2643Surprisingly, learning phrases longer than three words and learning phrases fromhigh-accuracy word-level alignment models does not have a strong impact on performance.
other,2-5-N03-1017,ak
impact on performance . Learning only
<term>
syntactically motivated phrases
</term>
degrades the performance of our systems
#2658Learning onlysyntactically motivated phrases degrades the performance of our systems.