other,17-3-J05-1003,bq |
additional
<term>
features
</term>
of the
<term>
|
tree
|
</term>
as evidence . The strength of our
|
#8706
A second model then attempts to improve upon this initial ranking, using additional features of thetree as evidence. |
tech,1-7-J05-1003,bq |
Street Journal treebank
</term>
. The
<term>
|
method
|
</term>
combined the
<term>
log-likelihood
</term>
|
#8800
Themethod 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. |
other,16-5-J05-1003,bq |
the
<term>
boosting approach
</term>
to
<term>
|
ranking problems
|
</term>
described in
<term>
Freund et al. (
|
#8775
We introduce a new method for the reranking task, based on the boosting approach toranking problems described in Freund et al. (1998). |
measure(ment),14-8-J05-1003,bq |
</term>
, a 13 % relative decrease in
<term>
|
F-measure
|
</term>
error over the
<term>
baseline model
|
#8849
The new model achieved 89.75% F-measure, a 13% relative decrease inF-measure error over the baseline model’s score of 88.2%. |
other,7-2-J05-1003,bq |
<term>
parser
</term>
produces a set of
<term>
|
candidate parses
|
</term>
for each input
<term>
sentence
</term>
|
#8670
The base parser produces a set ofcandidate parses for each input sentence, with associated probabilities that define an initial ranking of these parses. |
tech,6-9-J05-1003,bq |
The article also introduces a new
<term>
|
algorithm
|
</term>
for the
<term>
boosting approach
</term>
|
#8867
The article also introduces a newalgorithm for the boosting approach which takes advantage of the sparsity of the feature space in the parsing data. |
tech,4-5-J05-1003,bq |
</term>
into account . We introduce a new
<term>
|
method
|
</term>
for the
<term>
reranking task
</term>
|
#8763
We introduce a newmethod for the reranking task, based on the boosting approach to ranking problems described in Freund et al. (1998). |
other,16-9-J05-1003,bq |
</term>
which takes advantage of the
<term>
|
sparsity of the feature space
|
</term>
in the
<term>
parsing data
</term>
.
|
#8877
The article also introduces a new algorithm for the boosting approach which takes advantage of thesparsity of the feature space in the parsing data. |
other,26-4-J05-1003,bq |
, without concerns about how these
<term>
|
features
|
</term>
interact or overlap and without the
|
#8736
The strength of our approach is that it allows a tree to be represented as an arbitrary set of features, without concerns about how thesefeatures interact or overlap and without the need to define a derivation or a generative model which takes these features into account. |
other,45-4-J05-1003,bq |
generative model
</term>
which takes these
<term>
|
features
|
</term>
into account . We introduce a new
|
#8755
The strength of our approach is that it allows a tree to be represented as an arbitrary set of features, without concerns about how these features interact or overlap and without the need to define a derivation or a generative model which takes thesefeatures into account. |
tech,4-4-J05-1003,bq |
</term>
as evidence . The strength of our
<term>
|
approach
|
</term>
is that it allows a
<term>
tree
</term>
|
#8714
The strength of ourapproach is that it allows a tree to be represented as an arbitrary set of features, without concerns about how these features interact or overlap and without the need to define a derivation or a generative model which takes these features into account. |
measure(ment),18-8-J05-1003,bq |
<term>
F-measure
</term>
error over the
<term>
|
baseline model ’s score
|
</term>
of 88.2 % . The article also introduces
|
#8853
The new model achieved 89.75% F-measure, a 13% relative decrease in F-measure error over thebaseline model ’s score of 88.2%. |
other,12-10-J05-1003,bq |
<term>
algorithm
</term>
over the obvious
<term>
|
implementation
|
</term>
of the
<term>
boosting approach
</term>
|
#8899
Experiments show significant efficiency gains for the new algorithm over the obviousimplementation of the boosting approach. |
other,14-3-J05-1003,bq |
<term>
ranking
</term>
, using additional
<term>
|
features
|
</term>
of the
<term>
tree
</term>
as evidence
|
#8703
A second model then attempts to improve upon this initial ranking, using additionalfeatures of the tree as evidence. |
other,21-2-J05-1003,bq |
probabilities
</term>
that define an initial
<term>
|
ranking
|
</term>
of these
<term>
parses
</term>
. A second
|
#8684
The base parser produces a set of candidate parses for each input sentence, with associated probabilities that define an initialranking of these parses. |
other,37-4-J05-1003,bq |
overlap and without the need to define a
<term>
|
derivation
|
</term>
or a
<term>
generative model
</term>
|
#8747
The strength of our approach is that it allows a tree to be represented as an arbitrary set of features, without concerns about how these features interact or overlap and without the need to define aderivation or a generative model which takes these features into account. |
other,23-7-J05-1003,bq |
evidence from an additional 500,000
<term>
|
features
|
</term>
over
<term>
parse trees
</term>
that
|
#8822
The method combined the log-likelihood under a baseline model (that of Collins [1999]) with evidence from an additional 500,000features over parse trees that were not included in the original model. |
tech,8-10-J05-1003,bq |
significant efficiency gains for the new
<term>
|
algorithm
|
</term>
over the obvious
<term>
implementation
|
#8895
Experiments show significant efficiency gains for the newalgorithm over the obvious implementation of the boosting approach. |
tech,2-2-J05-1003,bq |
probabilistic parser
</term>
. The base
<term>
|
parser
|
</term>
produces a set of
<term>
candidate
|
#8665
The baseparser produces a set of candidate parses for each input sentence, with associated probabilities that define an initial ranking of these parses. |
tech,16-12-J05-1003,bq |
language parsing ( NLP )
</term>
, the
<term>
|
approach
|
</term>
should be applicable to many other
|
#8952
Although the experiments in this article are on natural language parsing (NLP), theapproach should be applicable to many other NLP problems which are naturally framed as ranking tasks, for example, speech recognition, machine translation, or natural language generation. |