#8137We introduce a new method for the reranking task, based on theboosting approach to ranking problems described in Freund et al. (1998).
other,7-2-J05-1003,ak
base parser
</term>
produces a set of
<term>
candidate parses
</term>
for each
<term>
input sentence
</term>
#8035The 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,ak
The article also introduces a new
<term>
algorithm
</term>
for the
<term>
boosting approach
</term>
#8232The article also introduces a newalgorithm for the boosting approach which takes advantage of the sparsity of the feature space in the parsing data.
other,19-9-J05-1003,ak
of the
<term>
sparsity
</term>
of the
<term>
feature space
</term>
in the
<term>
parsing data
</term>
.
#8245The article also introduces a new algorithm for the boosting approach which takes advantage of the sparsity of thefeature space in the parsing data.
other,26-4-J05-1003,ak
, without concerns about how these
<term>
features
</term>
interact or overlap and without the
#8101The 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,ak
generative model
</term>
which takes these
<term>
features
</term>
into account . We introduce a new
#8120The 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.
other,16-9-J05-1003,ak
</term>
which takes advantage of the
<term>
sparsity
</term>
of the
<term>
feature space
</term>
in
#8242The article also introduces a new algorithm for the boosting approach which takes advantage of thesparsity of the feature space in the parsing data.
other,14-3-J05-1003,ak
<term>
ranking
</term>
, using additional
<term>
features
</term>
of the
<term>
tree
</term>
as evidence
#8068A second model then attempts to improve upon this initial ranking, using additionalfeatures of the tree as evidence.
other,21-2-J05-1003,ak
probabilities
</term>
that define an initial
<term>
ranking
</term>
of these
<term>
parses
</term>
. A second
#8049The 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,ak
overlap and without the need to define a
<term>
derivation
</term>
or a
<term>
generative model
</term>
#8112The 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,ak
evidence from an additional 500,000
<term>
features
</term>
over
<term>
parse trees
</term>
that
#8187The 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.
measure(ment),14-8-J05-1003,ak
</term>
, a 13 % relative decrease in
<term>
F-measure error
</term>
over the
<term>
baseline model ’s
</term>
#8214The new model achieved 89.75% F-measure, a 13% relative decrease inF-measure error over the baseline model’s score of 88.2%.
tech,8-10-J05-1003,ak
significant efficiency gains for the new
<term>
algorithm
</term>
over the obvious implementation of
#8260Experiments show significant efficiency gains for the newalgorithm over the obvious implementation of the boosting approach.
tech,1-2-J05-1003,ak
<term>
probabilistic parser
</term>
. The
<term>
base parser
</term>
produces a set of
<term>
candidate
#8029Thebase parser produces a set of candidate parses for each input sentence, with associated probabilities that define an initial ranking of these parses.
model,18-8-J05-1003,ak
<term>
F-measure error
</term>
over the
<term>
baseline model ’s
</term>
score of 88.2 % . The article also
#8218The new model achieved 89.75% F-measure, a 13% relative decrease in F-measure error over thebaseline model ’s score of 88.2%.
other,16-2-J05-1003,ak
input sentence
</term>
, with associated
<term>
probabilities
</term>
that define an initial
<term>
ranking
#8044The base parser produces a set of candidate parses for each input sentence, with associatedprobabilities that define an initial ranking of these parses.
other,25-7-J05-1003,ak
additional 500,000
<term>
features
</term>
over
<term>
parse trees
</term>
that were not included in the original
#8189The method combined the log-likelihood under a baseline model (that of Collins [1999]) with evidence from an additional 500,000 features overparse trees that were not included in the original model.
model,2-3-J05-1003,ak
these
<term>
parses
</term>
. A second
<term>
model
</term>
then attempts to improve upon this
#8056A secondmodel then attempts to improve upon this initial ranking, using additional features of the tree as evidence.
other,10-4-J05-1003,ak
of our approach is that it allows a
<term>
tree
</term>
to be represented as an arbitrary
#8085The strength of our approach is that it allows atree 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.
tech,3-6-J05-1003,ak
Freund et al. ( 1998 ) . We apply the
<term>
boosting method
</term>
to parsing the
<term>
Wall Street Journal
#8154We apply theboosting method to parsing the Wall Street Journal treebank.