coreference data 0.0026784
algorithm coreference 0.002519692
coreference rules 0.002457503
same coreference 0.002415361
coreference resolution 0.0023854740000000003
pairwise coreference 0.0022311650000000002
coreference system 0.002192777
coreference decisions 0.002132103
coreference classifier 0.0021167
coreference knowledge 0.0020913200000000002
coreference relation 0.002077613
coreference systems 0.002077224
coreference classifiers 0.0020708790000000003
phrase coreference 0.002027652
coreference links 0.00198497
few coreference 0.00198053
coreference chain 0.001975387
easy coreference 0.001969542
coreference clas 0.001958707
coreference ruleset 0.0019525760000000001
coreference par 0.001940041
coreference sys 0.001926246
original coreference 0.001920197
coreference chains 0.001917316
coreference relationships 0.001914546
term coreference 0.00191255
coreference rela 0.0019125190000000001
training data 0.001814259
resolution data 0.001720594
coreference 0.00167164
data set 0.0016453940000000001
clustering algorithm 0.001639086
feature set 0.0016360250000000002
learning algorithm 0.0016234069999999999
semantic class 0.0015783709999999999
resolution algorithm 0.001561886
muc data 0.001516673
class information 0.0015106809999999998
feature selection 0.001508475
training corpus 0.0015068680000000002
rule learning 0.001496585
system performance 0.001466757
first rule 0.001451452
training set 0.001446133
training instances 0.001432347
feature values 0.001399807
additional data 0.0013663640000000001
selection algorithm 0.001359136
feature description 0.001329625
training instance 0.001316097
type feature 0.001314381
feature type 0.001314381
clustering task 0.001309127
data sets 0.001301689
algorithm ing 0.001297552
ing algorithm 0.001297552
learning system 0.001296492
preceding nps 0.001296081
learning task 0.001293448
performance gains 0.001287926
coreferent nps 0.0012662889999999999
ing clustering 0.001240534
performance improvements 0.001236453
resolution task 0.0012319269999999998
overall performance 0.001221961
ent nps 0.00122127
pruning algorithm 0.0012085309999999999
tem performance 0.001207572
relational features 0.001194993
hurt performance 0.001192815
class dis 0.001189122
class distributions 0.0011796649999999999
skewed class 0.0011778869999999999
training texts 0.001173795
same string 0.00116579
positive rules 0.001165424
training examples 0.001162074
minority class 0.001157469
class errors 0.001156441
different approach 0.001154132
classification rule 0.001152007
training document 0.001138632
evaluation corpus 0.001136065
new rules 0.001131438
anaphora resolution 0.001130162
learning framework 0.00112695
pronoun resolution 0.0011130649999999999
algorithm attempts 0.00110785
clustering level 0.001101637
positive rule 0.001100791
available training 0.001093737
number information 0.0010877959999999998
machine learning 0.001082531
rule pruning 0.001081709
test instance 0.001076885
link clustering 0.001074605
ing instances 0.001074348
hard training 0.001066239
pruning corpus 0.001059848
negative instances 0.0010504989999999999
