system text 0.002292458
semantic input 0.0019546349999999997
sentence input 0.001941207
output text 0.001903964
new text 0.001897368
first text 0.0018805699999999998
generation system 0.001854872
text number 0.001737668
surface text 0.001666454
language generation 0.001649323
final text 0.0016400949999999998
small text 0.001634384
data sets 0.001628607
short text 0.001617185
data flow 0.001577867
data sparseness 0.001573353
generation task 0.001537198
input figure 0.001501493
different texts 0.001469381
score input 0.00142261
input target 0.001406371
such features 0.0013937139999999999
single input 0.001369462
generation process 0.0013667800000000002
initial input 0.0013345499999999999
nlg system 0.001317035
new approach 0.0013108920000000001
typical input 0.001296419
input networks 0.0012803369999999999
points input 0.001273422
whole input 0.001269577
present system 0.001263935
input constructor 0.001253857
input repre 0.0012502709999999998
input specification 0.001248374
input specifica 0.00124176
usual input 0.00124176
semantic network 0.001231285
generation approaches 0.001230396
generation space 0.00122224
network sentence 0.001217857
riches system 0.001210956
syntactic features 0.001186695
semantic patterns 0.001184329
semantic extensions 0.001178494
different choice 0.001170582
linguistic features 0.001161958
possible output 0.0011559349999999999
guage generation 0.0011525110000000002
english language 0.001151736
natural language 0.00111206
free generation 0.001110372
overall approach 0.001108621
language variation 0.00109677
medicine information 0.001091879
other control 0.0010676840000000002
short texts 0.0010649969999999998
sible texts 0.0010640909999999999
other components 0.001063251
particular features 0.001057987
features factor 0.0010498460000000001
first choice 0.001029583
surface feature 0.001024826
formation texts 0.00101892
input 0.00100925
language varia 0.001006374
approach our 9.897320000000001E-4
possible decisions 9.889999999999999E-4
ing generator 9.82974E-4
statistical analysis 9.82011E-4
regression model 9.81489E-4
rule figure 9.80506E-4
system 9.77928E-4
eral approach 9.687680000000001E-4
tion model 9.671740000000001E-4
search algorithm 9.65802E-4
modular approach 9.64442E-4
different styles 9.624340000000001E-4
feature characteristics 9.619849999999999E-4
generator choice 9.575359999999999E-4
first type 9.48775E-4
consistent corpus 9.446459999999999E-4
such architectures 9.43943E-4
tic features 9.35455E-4
whole corpus 9.34245E-4
sentence 9.31957E-4
statistical correlation 9.28028E-4
other samples 9.276970000000001E-4
face features 9.2177E-4
dividual features 9.190800000000001E-4
model correlation 9.18472E-4
statistical results 9.17397E-4
generator decisions 9.16492E-4
sentences choice 8.97396E-4
possible combinations 8.89851E-4
regression scores 8.89427E-4
large number 8.85924E-4
first part 8.85539E-4
human sub 8.849610000000001E-4
generator control 8.8481E-4
