#28907An initial experiment, on a sample of 14 verbs which exhibit multiple complementation patterns, demonstrates that the technique achievesaccuracy comparable to previous approaches, which are all limited to a highly restricted set of subcategorization classes.
measure(ment),13-4-A97-1052,ak
the
<term>
system
</term>
improves the
<term>
accuracy
</term>
of a
<term>
parser
</term>
by an appreciable
#28939We also demonstrate that a subcategorization dictionary built with the system improves theaccuracy of a parser by an appreciable amount.
other,13-3-A97-1052,ak
verbs
</term>
which exhibit multiple
<term>
complementation patterns
</term>
, demonstrates that the
<term>
technique
#28899An initial experiment, on a sample of 14 verbs which exhibit multiplecomplementation patterns, demonstrates that the technique achieves accuracy comparable to previous approaches, which are all limited to a highly restricted set of subcategorization classes.
model,1-2-A97-1052,ak
<term>
textual corpora
</term>
. Each
<term>
dictionary entry
</term>
encodes the
<term>
relative frequency
#28868Eachdictionary entry encodes the relative frequency of occurrence of a comprehensive set of subcategorization classes for English.
other,17-2-A97-1052,ak
subcategorization classes
</term>
for
<term>
English
</term>
. An initial experiment , on a sample
#28884Each dictionary entry encodes the relative frequency of occurrence of a comprehensive set of subcategorization classes forEnglish.
tech,16-4-A97-1052,ak
improves the
<term>
accuracy
</term>
of a
<term>
parser
</term>
by an appreciable amount . Valiant
#28942We also demonstrate that a subcategorization dictionary built with the system improves the accuracy of aparser by an appreciable amount.
other,5-2-A97-1052,ak
dictionary entry
</term>
encodes the
<term>
relative frequency of occurrence
</term>
of a comprehensive set of
<term>
subcategorization
#28872Each dictionary entry encodes therelative frequency of occurrence of a comprehensive set of subcategorization classes for English.
other,14-2-A97-1052,ak
occurrence
</term>
of a comprehensive set of
<term>
subcategorization classes
</term>
for
<term>
English
</term>
. An initial
#28881Each dictionary entry encodes the relative frequency of occurrence of a comprehensive set ofsubcategorization classes for English.
other,37-3-A97-1052,ak
limited to a highly restricted set of
<term>
subcategorization classes
</term>
. We also demonstrate that a
<term>
#28923An initial experiment, on a sample of 14 verbs which exhibit multiple complementation patterns, demonstrates that the technique achieves accuracy comparable to previous approaches, which are all limited to a highly restricted set ofsubcategorization classes.
lr,5-4-A97-1052,ak
</term>
. We also demonstrate that a
<term>
subcategorization dictionary
</term>
built with the
<term>
system
</term>
#28931We also demonstrate that asubcategorization dictionary built with the system improves the accuracy of a parser by an appreciable amount.
tech,10-4-A97-1052,ak
subcategorization dictionary
</term>
built with the
<term>
system
</term>
improves the
<term>
accuracy
</term>
#28936We also demonstrate that a subcategorization dictionary built with thesystem improves the accuracy of a parser by an appreciable amount.
tech,7-1-A97-1052,ak
a novel technique and implemented
<term>
system for constructing a subcategorization dictionary
</term>
from
<term>
textual corpora
</term>
.
#28857We describe a novel technique and implementedsystem for constructing a subcategorization dictionary from textual corpora.
#28905An initial experiment, on a sample of 14 verbs which exhibit multiple complementation patterns, demonstrates that thetechnique achieves accuracy comparable to previous approaches, which are all limited to a highly restricted set of subcategorization classes.
lr,14-1-A97-1052,ak
subcategorization dictionary
</term>
from
<term>
textual corpora
</term>
. Each
<term>
dictionary entry
</term>
#28864We describe a novel technique and implemented system for constructing a subcategorization dictionary fromtextual corpora.