pos features 0.002671207
parsing model 0.002627352
parsing system 0.002614984
same features 0.002593606
special features 0.002428518
position features 0.002407073
pair features 0.0024069060000000003
individual features 0.002395007
dependency tree 0.002392186
cial features 0.0023733870000000002
neighbor features 0.0023733870000000002
training data 0.00223499
dependency label 0.00219743
chinese parsing 0.0021438539999999997
dependency type 0.002133803
parsing the 0.0021217149999999997
dependency labels 0.00210719
features 0.00209266
dependency arcs 0.002084087
corresponding dependency 0.002083585
projective parsing 0.002076422
deterministic parsing 0.0020620639999999997
parsing method 0.002042945
pendency parsing 0.002015649
parsing algo 0.002014287
dency parsing 0.001997952
next parser 0.0019585180000000002
unlabeled parser 0.001911674
test data 0.001775459
dependency 0.00174131
parser action 0.0017231780000000001
parsing 0.0016893
chinese data 0.001658384
last word 0.0015459620000000001
data repre 0.001496816
ﬁrst word 0.001477024
ing model 0.001425239
head words 0.0014046380000000002
stage system 0.001385411
parser 0.00137319
different labels 0.0013672600000000001
svm model 0.0013212459999999999
evaluation results 0.0012676879999999999
phrase attachment 0.001151369
attachment score 0.001148893
experimental results 0.001137827
good performance 0.001132973
average performance 0.0011167780000000001
ofﬁcial results 0.001100714
error analysis 0.001094279
input sentence 0.001031178
training 0.00103116
beled attachment 0.001026173
projective approach 0.001025544
shared task 0.001017481
root labeler 0.001013524
the pos 0.001010962
head dep 0.001010305
unlabeled accuracy 0.001010137
error analyses 0.001002672
wrong head 9.96085E-4
projective structure 9.78266E-4
general error 9.76888E-4
error rate 9.67561E-4
dency tree 9.595280000000001E-4
sequence problem 9.396459999999999E-4
model 9.38052E-4
additional knowledge 9.369109999999999E-4
system 9.25684E-4
proposed approach 9.22516E-4
next left 9.2182E-4
labeled accuracy 9.18304E-4
good parameters 8.868890000000001E-4
multilingual track 8.8476E-4
right neighbors 8.82461E-4
adaptation track 8.81657E-4
task organizers 8.751270000000001E-4
second stage 8.60058E-4
original input 8.515129999999999E-4
correct label 8.47717E-4
processed tokens 8.32774E-4
labeling problem 8.21446E-4
results 8.18284E-4
sequence labeler 8.12102E-4
domain adaptation 8.07689E-4
sequence labeling 7.984559999999999E-4
the maltparser 7.98058E-4
the lemma 7.954920000000001E-4
token rightmost 7.93401E-4
lingual track 7.88602E-4
input string 7.87992E-4
token leftmost 7.82629E-4
derivation history 7.81973E-4
child node 7.80365E-4
cpostag sequence 7.783029999999999E-4
ﬁrst stage 7.70031E-4
projective depen 7.63215E-4
token immedi 7.6207E-4
the cpos 7.617640000000001E-4
beled accuracy 7.60854E-4
