word features 0.0026632799999999996
such features 0.00207905
data set 0.0018563860000000002
training data 0.001847357
similar features 0.00182334
lexical features 0.001804401
semantic features 0.001785329
additional features 0.001768284
test data 0.001760945
tional features 0.0017509779999999998
related features 0.001747844
orthographic features 0.0017421189999999999
syntactic features 0.001717414
features pre 0.001710035
effective features 0.0017068559999999999
linguistic features 0.001689473
advanced features 0.001688092
conventional features 0.001687179
lated features 0.001682861
task data 0.001665721
word class 0.001637946
word vector 0.0015881609999999998
current word 0.001509163
word level 0.0014893419999999998
features 0.0014765
labeled word 0.0014663319999999999
unlabeled data 0.0014613710000000002
data sets 0.0014367490000000002
annotated data 0.001434815
ner system 0.001429583
word clusters 0.001415015
testing data 0.0014047920000000002
feature set 0.0014045519999999999
meta data 0.0014015050000000002
data preparation 0.001383507
training set 0.001346963
crf model 0.00134034
test set 0.001260551
tweets information 0.0012581110000000001
common words 0.001253266
short text 0.001229477
different feature 0.0012289879999999999
tional news 0.0012107939999999998
training tweets 0.0012081230000000002
ner task 0.001201232
entity types 0.001188936
entity type 0.001181312
baseline system 0.001175403
news articles 0.001171817
web text 0.001147562
stop words 0.001132614
context information 0.001132184
feature vector 0.0011279369999999999
new training 0.001126104
name entity 0.001119481
informal text 0.001112151
classifier table 0.001105993
conditional model 0.001101531
feature matrix 0.001097415
various feature 0.001078041
our model 0.001062355
model our 0.001062355
domain training 0.001061708
formal text 0.001057457
model extracts 0.001043397
feature extraction 0.001041577
other tweet 0.001040559
current ner 0.001036284
text window 0.001032561
fine model 0.0010300679999999999
specific ner 0.001030067
beled text 0.001029171
text message 0.0010290339999999999
mal text 0.001028027
level ner 0.001016463
ner systems 0.00100867
crf algorithm 0.001006529
knn classifier 0.001001233
global information 9.99992E-4
results table 9.99667E-4
other methods 9.90527E-4
feature sets 9.849149999999998E-4
knn training 9.7905E-4
novel ner 9.7875E-4
stanford ner 9.722839999999999E-4
limited information 9.71031E-4
final system 9.52156E-4
ford ner 9.520479999999999E-4
system pre 9.49217E-4
contextual information 9.478469999999999E-4
performance improvement 9.477139999999999E-4
learning table 9.46373E-4
different types 9.41182E-4
ing method 9.35444E-4
insufficient information 9.31752E-4
overview ner 9.296899999999999E-4
biomedical ner 9.264189999999999E-4
overall performance 9.24116E-4
other classifiers 9.21963E-4
user names 9.20915E-4
