#644A language learning experiment showed that assessors can differentiate native from non-native language essays in less than 100 words.
. Our
<term>
algorithm
</term>
reported more
than
99 %
<term>
accuracy
</term>
in both
<term>
language
#1281Our algorithm reported more than 99% accuracy in both language identification and key prediction.
rating
</term>
on average is only 5 % worse
than
the
<term>
top human-ranked sentence plan
#1454We show that the trained SPR learns to select a sentence plan whose rating on average is only 5% worse than the top human-ranked sentence plan.
alternative markers
</term>
, which includes other (
than
) , such ( as ) , and besides . These
<term>
#1836This paper presents a formal analysis for a large class of words called alternative markers, which includes other ( than), such (as), and besides.
#2109We show that the trainable sentence planner performs better than the rule-based systems and the baselines, and as well as the hand-crafted system.
#2636Surprisingly, learning phrases longer than three words and learning phrases from high-accuracy word-level alignment models does not have a strong impact on performance.
</term>
with
<term>
in-degree
</term>
greater
than
one and
<term>
out-degree
</term>
greater than
#3185For our purposes, a hub is a node in a graph with in-degree greater than one and out-degree greater than one.
than one and
<term>
out-degree
</term>
greater
than
one . We create a
<term>
word-trie
</term>
#3190For our purposes, a hub is a node in a graph with in-degree greater than one and out-degree greater than one.
simpler set of
<term>
model parameters
</term>
than
similar
<term>
phrase-based models
</term>
#3413In this paper, we describe a phrase-based unigram model for statistical machine translation that uses a much simpler set of model parametersthan similar phrase-based models.
a more effective
<term>
CFG filter
</term>
than
that of
<term>
LTAG
</term>
. We also investigate
#5137We demonstrate that an approximation of HPSG produces a more effective CFG filterthan that of LTAG.
generalize naturally to NLP structures other
than
<term>
parse trees
</term>
. This paper presents
#5577Although our experiments are focused on parsing, the techniques described generalize naturally to NLP structures other than parse trees.
score
</term>
that is significantly higher
than
that of the
<term>
baseline
</term>
. Following
#5900In our experiments, the method achieves a TRDR score that is significantly higher than that of the baseline.
SMT models
</term>
to be significantly lower
than
that of all the dedicated
<term>
WSD models
#6477We present controlled experiments showing the WSD accuracy of current typical SMT models to be significantly lower than that of all the dedicated WSD models considered.
integrating some kind of information other
than
<term>
grammar
</term>
sensu stricto into the
#7842Moreover, some examples are given that underline the necessity of integrating some kind of information other than grammar sensu stricto into the treebank.
significantly higher
<term>
accuracy
</term>
than
a state-of-the-art
<term>
coherence model
#8663Our experiments demonstrate that the induced model achieves significantly higher accuracythan a state-of-the-art coherence model.
simultaneously using less
<term>
memory
</term>
than
is required by current
<term>
decoder implementations
#8787In this paper we describe a novel data structure for phrase-based statistical machine translation which allows for the retrieval of arbitrarily long phrases while simultaneously using less memorythan is required by current decoder implementations.
labelled bracket F-score
</term>
of 76.2 , higher
than
previously reported results on the
<term>
#8976In this paper, we present an unlexicalized parser for German which employs smoothing and suffix analysis to achieve a labelled bracket F-score of 76.2, higher than previously reported results on the NEGRA corpus.
#9237Using a state-of-the-art Chinese word sense disambiguation model to choose translation candidates for a typical IBM statistical MT system, we find that word sense disambiguation does not yield significantly better translation qualitythan the statistical machine translation system alone.
</term>
yields a lower
<term>
error rate
</term>
than
the
<term>
HMM and Maxent models
</term>
on
#9544In general, our CRF model yields a lower error ratethan the HMM and Maxent models on the NIST sentence boundary detection task in speech, although it is interesting to note that the best results are achieved by three-way voting among the classifiers.
</term>
which is an order of magnitude smaller
than
<term>
Penn WSJ
</term>
. We present a
<term>
#10562We explored possible ways to obtain a compact lexicon, consistent with CCG principles, from a treebank which is an order of magnitude smaller than Penn WSJ.