D13-1073 |
hybrid approach , augmenting a
|
machine learner
|
to a set of handwritten rules
|
H05-1046 |
tagged data was used to train the
|
machine learner
|
. We compared this method with
|
E14-4022 |
incorporate the two feature spaces in a
|
machine learner
|
. 6 Acknowledgements This work
|
D12-1053 |
van den Bosch , 2005 ) . This
|
machine learner
|
is originally designed for part-of-speech
|
C04-1035 |
algorithms on these data sets . The rst
|
machine learner
|
used , SLIPPER , extracts optimised
|
I05-3003 |
noted that we use a different
|
machine learner
|
from the original method ( Nivre
|
I05-3003 |
leaves the decision to another
|
machine learner
|
that makes use of global fea
|
C04-1035 |
2003 ) . As with all memory-based
|
machine learners
|
, TiMBL stores representations
|
C02-1088 |
also made in other languages . As
|
machine learner
|
requires training and test examples
|
I05-3003 |
Similarly to Isozaki 's work , we use
|
machine learner
|
( SVMs ) to construct the root
|
D15-1035 |
the loss function of a POS-tag
|
machine learner
|
, resulting in improved performance
|
D08-1004 |
annotator can provide hints to a
|
machine learner
|
by highlighting contextual "
|
I05-3003 |
and then gives these features to
|
machine learner
|
. The machine learner uses the
|
D12-1053 |
memory-based system except that no
|
machine learner
|
is included . The most frequent
|
D10-1084 |
effectiveness of various features and
|
machine learners
|
for this task . While a simple
|
H05-1046 |
tagged data was used to train a
|
machine learner
|
, which disambiguated toponyms
|
E09-1070 |
Wikipedia-based feature in their
|
machine learner
|
. Such approaches are limited
|
J08-3001 |
disagreements add patterns for the
|
machine learner
|
to find . Unfortunately , the
|
D12-1064 |
employ a maximum entropy ( MaxEnt )
|
machine learner
|
-- MegaM ( fifth re - lease )
|
C02-1088 |
individual result . In our module ,
|
machine learners
|
train with the training examples
|