context word 0.002876715
word context 0.002876715
word node 0.0026864920000000004
word nodes 0.002509013
left word 0.002438574
right word 0.002418622
text corpus 0.001829858
context words 0.001809345
abbreviation corpus 0.001635161
abbreviation candidates 0.001590158
training data 0.001578272
context features 0.001567999
abbreviation candidate 0.001535119
large text 0.001526649
abbreviation task 0.0015118150000000001
character similarity 0.0014426909999999999
chinese abbreviation 0.001394183
contextual graph 0.001383022
news text 0.0013823499999999999
walk method 0.001361175
large corpus 0.001336467
markov model 0.001295889
natural text 0.001292432
context information 0.001285314
ranking model 0.00127875
bipartite graph 0.001273452
data structures 0.0012728499999999998
text cor 0.00126707
new abbreviation 0.001253332
level features 0.001241798
abbreviation generation 0.0012407569999999999
abbreviation pairs 0.00123552
corresponding abbreviation 0.001228858
graph representation 0.0012201269999999999
news corpus 0.001192168
graph construction 0.0011659819999999999
node set 0.0011634599999999998
graph rep 0.001161451
complex data 0.001157605
graph randomwalk 0.0011547139999999998
candidate vector 0.001151413
candidate generation 0.00114523
character context 0.001140626
generation task 0.001121926
full form 0.001121335
force method 0.001113006
similar context 0.0011076929999999999
words 0.00110589
stage method 0.001101976
ranking method 0.001098385
improved method 0.00107958
machine translation 0.001075727
candidate abbreviations 0.001069135
date abbreviation 0.001066493
abbreviation prediction 0.0010539199999999999
relevant candidates 0.001042147
context nodes 0.001039208
form source 0.001036114
system chinese 0.001035888
contextual information 0.001034144
breviation candidates 0.001023672
model 0.00102268
first node 0.001013836
similarity 0.00100552
good results 9.72795E-4
edge weight 9.51527E-4
ilar candidate 9.49158E-4
candidate genera 9.46999E-4
context window 9.46068E-4
prediction task 9.350890000000001E-4
graph 9.30737E-4
chinese abbreviations 9.28199E-4
random walk 8.90931E-4
form vfull 8.888790000000001E-4
source node 8.85404E-4
character sequence 8.72649E-4
competitive results 8.709E-4
acceptable results 8.709E-4
sequence labeling 8.68349E-4
features 8.64544E-4
first case 8.59339E-4
previous work 8.57932E-4
transition probability 8.47797E-4
other hand 8.465199999999999E-4
method 8.42315E-4
similar character 8.41409E-4
same concept 8.360449999999999E-4
information retrieval 8.359730000000001E-4
language processing 8.27001E-4
corpus 8.19838E-4
abbreviation 8.15323E-4
construction algorithm 8.06597E-4
full forms 8.06532E-4
probability pij 8.035E-4
final node 8.02968E-4
large amount 7.992730000000001E-4
labeling problem 7.95628E-4
news texts 7.93052E-4
new abbreviations 7.87348E-4
translation 7.82314E-4
