D09-1116 |
using off-the-shelf POS taggers or
|
PCFG parsers
|
. However , the amount of information
|
D12-1095 |
extracted from the chart of a base
|
PCFG parser
|
, usually employing a heavy pruning
|
D09-1087 |
improvements from self-training a
|
PCFG parser
|
on the standard WSJ training
|
D14-1100 |
classifier outperforms the current
|
PCFG parser
|
. Furthermore , it can be easily
|
C04-1104 |
application of the acquired lexicon to a
|
PCFG parser
|
indicates great potentialities
|
C04-1204 |
on other formalisms including
|
PCFG parsers
|
. At the same time , following
|
D14-1100 |
the same as the result of our
|
PCFG parser
|
because both systems use contextual
|
D11-1064 |
Since we do not have an a able
|
PCFG parser
|
as in ( Huang and Harper , 2009
|
C04-1204 |
still below the state-of-theart
|
PCFG parsers
|
( Collins , 1999 ; Charniak ,
|
D14-1100 |
effectively integrated into the
|
PCFG parser
|
and general statistical parsing
|
D12-1096 |
parser ( - F ) , which combines the
|
PCFG parser
|
with a lexicalised dependency
|
D12-1095 |
Details We used the first-stage
|
PCFG parser
|
of Charniak and Johnson ( 2005
|
D14-1098 |
trained on Penn Treebank data . The
|
PCFG parser
|
is generalized to take the lattice
|
D09-1015 |
labeli_1 + labeli chart-based
|
PCFG parser
|
, except that instead of putting
|
D10-1050 |
stand-alone sentence , if the
|
PCFG parser
|
has marked i with an S ( sen
|
D12-1095 |
two parsers are state-of-the-art
|
PCFG parsers
|
for English and German , respectively
|
C04-1204 |
those parsers was still below
|
PCFG parsers
|
( Collins , 1999 ; Charniak ,
|
C04-1204 |
analysis has not outperformed
|
PCFG parsers
|
in terms of the accuracy of surface
|
D09-1087 |
Charniak 's parser is a lexicalized
|
PCFG parser
|
that models lexicalized dependencies
|
D12-1096 |
consider both the unlexicalised
|
PCFG parser
|
( - U ) and the factored parser
|