new parser 0.002382831
standard parser 0.002248001
parser accuracy 0.0022112069999999998
parsing model 0.002075066
parser output 0.002066465
account parser 0.002036076
parser confidence 0.00203368
art parser 0.002021367
parser performances 0.00201273
lexical information 0.001898388
lexical dependencies 0.0018966819999999998
parsing results 0.001832113
lexical resource 0.001797169
parser 0.00179602
lexical counts 0.0017741609999999998
lexical affinity 0.001771917
different features 0.001759051
lexical values 0.0017427229999999998
lexical phenomena 0.001740218
computing lexical 0.001719743
lexical informa 0.0017192589999999999
model word 0.0017152320000000001
lexical affinities 0.0016947969999999998
lexical sparsity 0.0016841039999999999
parsing method 0.001659093
second parsing 0.001620682
such data 0.001496818
data set 0.001493184
parsing stage 0.001490768
different results 0.001490333
double parsing 0.001481586
lexical 0.00146785
new features 0.001429612
parse tree 0.001377942
new parse 0.001375842
parse set 0.001352265
different number 0.001346448
other features 0.001331753
dependency trees 0.0013131
parse trees 0.001273175
training corpus 0.0012700250000000001
parsing 0.00125803
development data 0.0012501600000000002
accuracy dependency 0.001244143
new error 0.0012104120000000001
different configurations 0.001209397
the data 0.001204828
data sparseness 0.001201186
different values 0.001191123
single parse 0.001179073
subject dependency 0.0011770370000000001
different corpora 0.0011753179999999998
first pos 0.001169098
word occurrences 0.001159526
dependency types 0.0011508760000000001
training treebank 0.00114805
filtering data 0.001147591
different scenarios 0.001147071
different kinds 0.0011409
different instantiations 0.00114043
different genres 0.001135034
different sce 0.001133865
correct parse 0.001129068
optimal parse 0.00112795
word clus 0.001116132
order model 0.001111694
relative error 0.001110245
probabilistic parsers 0.001096531
same number 0.001092967
pos tag 0.001075105
error type 0.001055561
error number 0.001053799
french corpus 0.001047694
dependency freq 0.0010453230000000001
training approach 0.001038216
training process 0.001018729
new information 0.001017349
reference parse 0.001016062
same size 0.001015122
parse hypothesis 0.001014633
pos tags 0.001001411
ing results 9.97866E-4
same way 9.87374E-4
gold pos 9.75697E-4
evaluation corpus 9.74899E-4
error rate 9.59795E-4
new errors 9.50888E-4
ftb test 9.489869999999999E-4
global error 9.464639999999999E-4
error types 9.45521E-4
common error 9.38202E-4
raw corpus 9.34254E-4
french treebank 9.25719E-4
other dependencies 9.177839999999999E-4
interesting results 9.17602E-4
open pos 9.015850000000001E-4
new strategy 8.991030000000001E-4
error ratio 8.97991E-4
such techniques 8.94606E-4
new parses 8.92711E-4
