#2189We report on different aspects of the predictive performance of our models, including the influence of various training and testing factors on predictive performance, and examine the relationships among the target variables.
lr,15-1-N03-1001,ak
<term>
manual transcription
</term>
of
<term>
training
data
</term>
. The method combines
<term>
domain
#2221This paper describes a method for utterance classification that does not require manual transcription oftraining data.
tech,4-3-N03-1001,ak
</term>
. In our method ,
<term>
unsupervised
training
</term>
is first used to train a
<term>
phone
#2261In our method, unsupervised training is first used to train a phone n-gram model for a particular domain; the output of recognition with this model is then passed to a phone-string classifier.
lr,2-1-N03-2003,ak
tagging result
</term>
. Sources of
<term>
training
data
</term>
suitable for
<term>
language modeling
#3017Sources oftraining data suitable for language modeling of conversational speech are limited.
lr,7-2-N03-2003,ak
limited . In this paper , we show how
<term>
training
data
</term>
can be supplemented with
<term>
#3036In this paper, we show howtraining data can be supplemented with text from the web filtered to match the style and/or topic of the target recognition task, but also that it is possible to get bigger performance gains from the data by using class-dependent interpolation of N-grams.
bootstrapping procedure
</term>
is implemented as
training
two
<term>
successive learners
</term>
. First
#3341The bootstrapping procedure is implemented as training two successive learners.
tech,1-4-N03-2036,ak
trigram language model
</term>
. During
<term>
training
</term>
, the
<term>
blocks
</term>
are learned
#3446Duringtraining, the blocks are learned from source interval projections using an underlying word alignment.
lr,27-2-P03-1050,ak
parallel corpus
</term>
as its sole
<term>
training
resources
</term>
. No
<term>
parallel text
#4475The stemming model is based on statistical machine translation and it uses an English stemmer and a small (10K sentences) parallel corpus as its soletraining resources.
other,7-3-P03-1050,ak
parallel text
</term>
is needed after the
<term>
training
phase
</term>
.
<term>
Monolingual , unannotated
#4485No parallel text is needed after thetraining phase.
lr,34-5-P03-1051,ak
expanded
<term>
vocabulary
</term>
and
<term>
training
corpus
</term>
. The resulting
<term>
Arabic
#4742To improve the segmentation accuracy, we use an unsupervised algorithm for automatically acquiring new stems from a 155 million word unsegmented corpus, and re-estimate the model parameters with the expanded vocabulary andtraining corpus.
lr,11-2-P03-1058,ak
automatically acquire
<term>
sense-tagged
training
data
</term>
from
<term>
English-Chinese parallel
#4835In this paper, we evaluate an approach to automatically acquire sense-tagged training data from English-Chinese parallel corpora, which are then used for disambiguating the nouns in the SENSEVAL-2 English lexical sample task.
lr,13-1-H05-1012,ak
Arabic-English
</term>
based on
<term>
supervised
training
data
</term>
. We demonstrate that it is
#5289This paper presents a maximum entropy word alignment algorithm for Arabic-English based on supervised training data.
lr,8-2-H05-1012,ak
demonstrate that it is feasible to create
<term>
training
material
</term>
for problems in
<term>
machine
#5300We demonstrate that it is feasible to createtraining material for problems in machine translation and that a mixture of supervised and unsupervised methods yields superior performance.
model,11-3-H05-1064,ak
assignments based on a
<term>
discriminative
training
criterion
</term>
.
<term>
Training
</term>
and
#5482The model learns to automatically make these assignments based on a discriminative training criterion.
tech,0-4-H05-1064,ak
discriminative training criterion
</term>
.
<term>
Training
</term>
and
<term>
decoding
</term>
with the
<term>
#5485The model learns to automatically make these assignments based on a discriminative training criterion.Training and decoding with the model requires summing over an exponential number of hidden-variable assignments: the required summations can be computed efficiently and exactly using dynamic programming.
tech,16-3-H05-1095,ak
<term>
phrases
</term>
, as well as a
<term>
training
method
</term>
based on the
<term>
maximization
#5630A statistical translation model is also presented that deals such phrases, as well as atraining method based on the maximization of translation accuracy, as measured with the NIST evaluation metric.
lr,17-5-H05-1095,ak
allows to better generalize from the
<term>
training
data
</term>
. This paper investigates some
#5675Experimental results are presented, that demonstrate how the proposed method allows to better generalize from thetraining data.
lr,15-1-P05-1046,ak
limited by the need for
<term>
supervised
training
data
</term>
. We demonstrate that for certain
#9032The applicability of many current information extraction techniques is severely limited by the need for supervised training data.
tech,8-1-P05-1069,ak
In this paper , we present a novel
<term>
training
method
</term>
for a
<term>
localized phrase-based
#9937In this paper, we present a noveltraining method for a localized phrase-based prediction model for statistical machine translation (SMT).