training algorithm 0.002362913
general algorithm 0.001996367
feature set 0.0019842289999999997
perceptron algorithm 0.001923561
such feature 0.001915932
estimation algorithm 0.0018504250000000002
algorithm updates 0.001833194
tron algorithm 0.001795532
viterbi algorithm 0.00177164
ceptron algorithm 0.001749952
generic algorithm 0.0017456820000000001
feature vector 0.0017412859999999999
training data 0.001676141
current model 0.0016315540000000001
training set 0.001610302
feature count 0.0015990129999999998
local feature 0.001590698
global feature 0.001585242
feature sets 0.001544537
feature type 0.001527723
parameter set 0.001513508
algorithm 0.00151183
feature vectors 0.001495317
generative model 0.001484707
entropy model 0.00147014
feature vec 0.0014598279999999998
feature templates 0.001457981
dimensional feature 0.001457981
pos tag 0.001426966
training sequence 0.001413417
tag sequence 0.001381982
training example 0.0013621459999999998
training error 0.001324714
training examples 0.001310009
test data 0.001286068
function gen 0.001285227
data examples 0.0012839840000000002
current word 0.001273852
parameter vector 0.001270565
tag task 0.00125935
current tag 0.001256782
arbitrary function 0.0012294390000000001
feature 0.00122501
other models 0.001220706
test set 0.001220229
results method 0.001212587
indicator function 0.001199109
model 0.00119442
current parameter 0.001191423
previous word 0.001191236
ing set 0.00118014
tag pair 0.001178413
previous tag 0.001174166
pos sequence 0.001169652
training sentence 0.001163274
learning algorithms 0.001158741
training iterations 0.00115594
parameter values 0.001152872
training samples 0.001148296
next word 0.001146246
data sets 0.001144585
trigram tag 0.001143657
tag trigram 0.001143657
word dis 0.001130895
training sample 0.001130474
next tag 0.001129176
additional set 0.001119972
entire tag 0.001110898
tag sequences 0.001092893
parameter estimation 0.0010928840000000001
tag symbols 0.001092813
sible tag 0.001090715
development data 0.0010860560000000002
theoretical results 0.001075257
rare word 0.0010744539999999999
tagging results 0.0010608240000000001
data exam 0.001059332
null tag 0.001057801
inseparable data 0.0010575670000000001
rect tag 0.001054251
parameter weights 0.001053689
method error 0.00105093
learning task 0.001044079
alternative parameter 0.001034699
pos tagging 0.0010328540000000002
parameter vectors 0.001024596
log probability 0.001020348
development set 0.001020217
parameter settings 0.001017975
only features 0.001012247
bayesian approach 0.001007399
various parameter 0.001007108
new algorithms 0.001003491
local features 0.001001052
tagging models 9.98165E-4
similar features 9.9602E-4
global features 9.95596E-4
other words 9.90987E-4
perceptron method 9.8903E-4
parameter estima 9.88854E-4
