|
probabilistic context-free grammars
</term>
working in
|
a
|
' synchronous ' way . Two
<term>
hardness
|
#7470
These models can be viewed as pairs of probabilistic context-free grammars working in a 'synchronous' way. |
|
Generation of Referring Expressions
</term>
: (
|
a
|
)
<term>
numeric-valued attributes
</term>
|
#10661
This paper discusses two problems that arise in the Generation of Referring Expressions: ( a) numeric-valued attributes, such as size or location; (b) perspective-taking in reference. |
|
</term>
achieved 89.75 %
<term>
F-measure
</term>
,
|
a
|
13 % relative decrease in
<term>
F-measure
|
#8843
The new model achieved 89.75% F-measure, a 13% relative decrease in F-measure error over the baseline model’s score of 88.2%. |
|
automatically acquiring new
<term>
stems
</term>
from
|
a
|
155 million
<term>
word
</term><term>
unsegmented
|
#4723
To 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 and training corpus. |
|
features
</term>
. The best system obtains
|
a
|
18.6 % improvement over the
<term>
baseline
|
#9640
The best system obtains a 18.6% improvement over the baseline on a standard Arabic-English translation task. |
|
questions correctly answered
</term>
, and
|
a
|
32.8 % improvement according to the
<term>
|
#2424
Experiments evaluating the effectiveness of our answer resolution algorithm show a 35.0% relative improvement over our baseline system in the number of questions correctly answered, and a 32.8% improvement according to the average precision metric. |
|
<term>
answer resolution algorithm
</term>
show
|
a
|
35.0 % relative improvement over our
<term>
|
#2406
Experiments evaluating the effectiveness of our answer resolution algorithm show a 35.0% relative improvement over our baseline system in the number of questions correctly answered, and a 32.8% improvement according to the average precision metric. |
|
accuracy
</term>
rate from 60 % to 75 % ,
|
a
|
37 % reduction in error . We discuss
<term>
|
#19045
In head-to-head tests against one of the best existing robust probabilistic parsing models, which we call P-CFG, the HBG model significantly outperforms P-CFG, increasing the parsing accuracy rate from 60% to 75%, a 37% reduction in error. |
|
<term>
error rate
</term>
dropped to 4.1 % ---
|
a
|
45 % reduction in
<term>
error
</term>
compared
|
#17206
Using only 40 utterances from the target speaker for adaptation, the error rate dropped to 4.1% --- a 45% reduction in error compared to the SI result. |
|
</term>
. The models were constructed using
|
a
|
5K
<term>
vocabulary
</term>
and trained using
|
#21247
The models were constructed using a 5K vocabulary and trained using a 76 million word Wall Street Journal text corpus. |
|
recognition system
</term>
, experiments show
|
a
|
7 % improvement in
<term>
recognition accuracy
|
#21271
Using the BU recognition system, experiments show a 7% improvement in recognition accuracy with the mixture trigram models as compared to using a trigram model. |
|
<term>
SI recognition
</term>
, we achieved
|
a
|
7.5 %
<term>
word error rate
</term>
on a standard
|
#17081
With only 12 training speakers for SI recognition, we achieved a 7.5% word error rate on a standard grammar and test set from the DARPA Resource Management corpus. |
|
<term>
vocabulary
</term>
and trained using
|
a
|
76 million
<term>
word
</term><term>
Wall Street
|
#21253
The models were constructed using a 5K vocabulary and trained using a 76 million word Wall Street Journal text corpus. |
|
, the resulting
<term>
tagger
</term>
gives
|
a
|
97.24 %
<term>
accuracy
</term>
on the
<term>
|
#2988
Using these ideas together, the resulting tagger gives a 97.24% accuracy on the Penn Treebank WSJ, an error reduction of 4.4% on the best previous single automatically learned tagging result. |
|
</term>
. The
<term>
model
</term>
is based on
|
a
|
<term>
balance matching operation
</term>
for
|
#19820
The model is based on a balance matching operation for two lists of the feature sets, which provides four effects: the reduction of analysis cost, the improvement of word disambiguation, the interpretation of ellipses, and robust analysis. |
|
combined the
<term>
log-likelihood
</term>
under
|
a
|
<term>
baseline model
</term>
( that of
<term>
|
#8805
The method combined the log-likelihood under a baseline model (that of Collins [1999]) with evidence from an additional 500,000 features over parse trees that were not included in the original model. |
|
as either coherent or incoherent ( given
|
a
|
<term>
baseline
</term>
of 54.55 % ) . We propose
|
#2532
An evaluation of our system against the annotated data shows that, it successfully classifies 73.2% in a German corpus of 2.284 SRHs as either coherent or incoherent (given a baseline of 54.55%). |
|
the system yields higher performance than
|
a
|
<term>
baseline
</term>
on all three aspects
|
#6747
Results indicate that the system yields higher performance than a baseline on all three aspects. |
|
contains a
<term>
recognition network
</term>
,
|
a
|
<term>
basic mapping
</term>
, additional
<term>
|
#12484
Each generalized metaphor contains a recognition network, a basic mapping, additional transfer mappings, and an implicit intention component. |
|
semantic constraints
</term>
and thus provide
|
a
|
basis for a useful
<term>
disambiguation
|
#16714
The results of the experiment show that in most of the cases the cooccurrence statistics indeed reflect the semantic constraints and thus provide a basis for a useful disambiguation tool. |