#397At MIT Lincoln Laboratory, we have been developing a Korean-to-English machine translation system CCLINC (Common Coalition Language System at Lincoln Laboratory).
tech,30-1-H01-1042,ak
</term>
, to the
<term>
output
</term>
of
<term>
machine
translation ( MT ) systems
</term>
. We believe
#574The purpose of this research is to test the efficacy of applying automated evaluation techniques, originally devised for the evaluation of human language learners, to the output ofmachine translation (MT) systems.
tech,24-2-H01-1042,ak
process
</term>
and the development of
<term>
machine
translation systems
</term>
. This , the
#605We believe that these evaluation techniques will provide information about both the human language learning process, the translation process and the development ofmachine translation systems.
other,16-6-H01-1042,ak
from duplicating the experiment using
<term>
machine
translation output
</term>
. Subjects were
#678We tested this to see if similar criteria could be elicited from duplicating the experiment usingmachine translation output.
other,11-8-H01-1042,ak
human translations
</term>
, others were
<term>
machine
translation outputs
</term>
. The subjects
#708Some of the extracts were expert human translations, others weremachine translation outputs.
other,24-9-H01-1042,ak
expert human translation
</term>
or a
<term>
machine
translation
</term>
. Additionally , they
#736The subjects were given three minutes per extract to determine whether they believed the sample output to be an expert human translation or amachine translation.
tech,28-4-H01-1055,ak
</term>
can be overcome by employing
<term>
machine
learning techniques
</term>
. In this paper
#1023We show how research in generation can be adapted to dialog systems, and how the high cost of hand-crafting knowledge-based generation systems can be overcome by employingmachine learning techniques.
tech,5-1-P01-1070,ak
</term>
. We describe a set of
<term>
supervised
machine
learning experiments
</term>
centering on
#2131We describe a set of supervised machine learning experiments centering on the construction of statistical models of WH-questions.
tech,8-1-N03-1004,ak
of
<term>
ensemble methods
</term>
in
<term>
machine
learning
</term>
and other areas of
<term>
#2315Motivated by the success of ensemble methods inmachine learning and other areas of natural language processing, we developed a multi-strategy and multi-source approach to question answering which is based on combining the results from different answering agents searching for answers in multiple corpora.
tech,11-1-N03-2036,ak
unigram model
</term>
for
<term>
statistical
machine
translation
</term>
that uses a much simpler
#3402In this paper, we describe a phrase-based unigram model for statistical machine translation that uses a much simpler set of model parameters than similar phrase-based models.
tech,6-2-P03-1050,ak
model
</term>
is based on
<term>
statistical
machine
translation
</term>
and it uses an
<term>
English
#4455The stemming model is based on statistical machine translation and it uses an English stemmer and a small (10K sentences) parallel corpus as its sole training resources.
tech,16-1-H05-1005,ak
input to correct
<term>
errors
</term>
in
<term>
machine
translation
</term>
and thus improve the
#5167In this paper, we use the information redundancy in multilingual input to correct errors inmachine translation and thus improve the quality of multilingual summaries.
tech,6-4-H05-1005,ak
documents . Further , the use of multiple
<term>
machine
translation systems
</term>
provides yet
#5228Further, the use of multiplemachine translation systems provides yet more redundancy, yielding different ways to realize that information in English.
other,6-5-H05-1005,ak
demonstrate how
<term>
errors
</term>
in the
<term>
machine
translations
</term>
of the input Arabic
#5252We demonstrate how errors in themachine translations of the input Arabic documents can be corrected by identifying and generating from such redundancy, focusing on noun phrases.
tech,13-2-H05-1012,ak
training material
</term>
for problems in
<term>
machine
translation
</term>
and that a mixture of
#5305We demonstrate that it is feasible to create training material for problems inmachine translation and that a mixture of supervised and unsupervised methods yields superior performance.
measure(ment),15-4-H05-1012,ak
as well as improvement on several
<term>
machine
translation tests
</term>
. Performance of
#5348Significant improvement over traditional word alignment techniques is shown as well as improvement on severalmachine translation tests.
tech,4-1-H05-1095,ak
presents a
<term>
phrase-based statistical
machine
translation method
</term>
, based on
<term>
#5587This paper presents a phrase-based statistical machine translation method, based on non-contiguous phrases, i.e. phrases with gaps.
tech,20-1-H05-1101,ak
been adopted in the literature on
<term>
machine
translation
</term>
. These
<term>
models
</term>
#5698This paper investigates some computational problems associated with probabilistic translation models that have recently been adopted in the literature onmachine translation.
tech,8-1-H05-1117,ak
<term>
automatic evaluation
</term>
of
<term>
machine
translation
</term>
and
<term>
document summarization
#5914Following recent developments in the automatic evaluation ofmachine translation and document summarization, we present a similar approach, implemented in a measure called POURPRE, for automatically evaluating answers to definition questions.
tech,12-2-H05-2007,ak
tool
</term>
intended for developers of
<term>
machine
translation systems
</term>
, and demonstrate
#6057We incorporate this analysis into a diagnostic tool intended for developers ofmachine translation systems, and demonstrate how our application can be used by developers to explore patterns in machine translation output.