training data 0.003126473
data set 0.002973478
unlabeled data 0.0029709429999999998
labeled data 0.002898861
same data 0.002835669
test data 0.002825813
syntactic data 0.002759672
wsj data 0.0027306879999999998
ner data 0.002719494
speech data 0.002691267
data sets 0.002673479
separate data 0.0026642669999999997
data points 0.002656487
single data 0.0026437049999999997
ment data 0.0026225009999999997
development data 0.002621708
conll data 0.002595928
dev data 0.0025949149999999997
data increases 0.00258788
ace data 0.002585531
training model 0.002055843
feature set 0.001778838
ner model 0.001648864
markov model 0.0016346660000000001
pos feature 0.001578087
model confidence 0.0015730540000000002
formal model 0.0015697060000000001
average model 0.0015694040000000002
own model 0.0015691840000000002
hints model 0.001568739
account model 0.0015194120000000001
mal model 0.001515559
different models 0.00150855
feature sets 0.001478839
learning problem 0.001471577
such learning 0.0014699189999999999
learning results 0.001469644
feature hints 0.001444729
discriminative learning 0.001393419
disjoint feature 0.001391977
constraint function 0.001390683
word alignment 0.00134771
model 0.00128402
pac learning 0.001266408
other words 0.001252019
such information 0.001220126
baseline performance 0.001204881
state models 0.001197787
new models 0.001186608
markov models 0.001183853
following models 0.001176112
syntactic information 0.001171349
different output 0.001169288
unlabeled corpus 0.0011634509999999998
labeled set 0.0011630389999999998
feature 0.00116001
constraint functions 0.0011591779999999999
labeling algorithm 0.001150664
separate models 0.001142824
syntactic features 0.001127152
information extraction 0.001124132
compatibility function 0.00108779
straint function 0.001066266
many training 0.001065811
same baseline 0.00105107
baseline systems 0.001046789
different approaches 0.001037175
mitchell algorithm 0.001033637
training setting 0.0010271540000000002
first label 0.001022206
gazetteer information 0.001021048
learning 0.00101612
other constraints 0.001009382
second set 0.001007912
baseline methods 0.001005726
labeled sentences 9.77973E-4
pos constraint 9.768519999999998E-4
unlabeled examples 9.43444E-4
possible output 9.39008E-4
ner baseline 9.34895E-4
discourse parsing 9.12601E-4
syntactic labels 9.12306E-4
labeled ner 9.09055E-4
target functions 9.0898E-4
similar problem 9.04573E-4
random subset 9.043040000000001E-4
other task 8.74341E-4
national corpus 8.74026E-4
shallow parsing 8.73944E-4
ner labels 8.72128E-4
labeled examples 8.71362E-4
language processing 8.70741E-4
compatibility functions 8.56285E-4
actual sentence 8.520539999999999E-4
large amount 8.50527E-4
unlabeled distri 8.47043E-4
baseline hmms 8.45378E-4
syntactic sentences 8.38784E-4
output space 8.36835E-4
random fields 8.354790000000001E-4
