language model 0.0015324079999999999
field model 0.0015121649999999998
probability model 0.0013870509999999998
model probability 0.0013870509999999998
different features 0.001344519
alignment model 0.0013414079999999999
unsupervised model 0.001324341
probabilistic model 0.0013134899999999998
feature vector 0.001298625
generative model 0.001297897
model parameters 0.001291121
feature weights 0.001278227
feature templates 0.0012586490000000001
generic feature 0.0012071200000000001
field set 0.001173055
training data 0.001130866
first word 0.001103166
many features 0.001092464
generation system 0.001083018
test set 0.001078916
word choice 0.001064522
generation task 0.001056122
model 0.00103573
following set 0.001027595
small set 0.001017057
additional features 0.001010995
generation template 0.00100141
feature 9.96834E-4
output text 9.80156E-4
set decisions 9.77702E-4
set decision 9.61111E-4
fluency semantic 9.56354E-4
semantic correctness 9.51348E-4
global features 9.50023E-4
same meaning 9.3636E-4
forecast generation 9.3516E-4
custom features 9.31808E-4
world state 9.22737E-4
ing field 9.142449999999999E-4
other words 9.06529E-4
generation templates 8.98977E-4
generation process 8.97958E-4
set deci 8.96176E-4
expressive set 8.95745E-4
full generation 8.948700000000001E-4
different decisions 8.92149E-4
time field 8.905219999999999E-4
wind weather 8.84594E-4
optimal text 8.83933E-4
dataset data 8.77458E-4
different domains 8.75853E-4
training examples 8.75382E-4
baseline system 8.714840000000001E-4
descriptive text 8.65786E-4
different times 8.5741E-4
unsupervised learning 8.535610000000001E-4
state con 8.4868E-4
field value 8.47745E-4
robust generation 8.4478E-4
learning problem 8.345620000000001E-4
powerful generation 8.34554E-4
annotated data 8.28197E-4
sumtime data 8.187380000000001E-4
first record 8.135589999999999E-4
record type 8.11641E-4
different realizations 8.1071E-4
supervised learning 7.997400000000001E-4
various grammar 7.96182E-4
training scenario 7.94494E-4
deterministic function 7.918900000000001E-4
baseline records 7.884319999999999E-4
fields fields 7.8127E-4
logical form 7.7808E-4
field values 7.76283E-4
first type 7.75034E-4
record choice 7.749149999999999E-4
random field 7.72399E-4
learning now 7.643960000000001E-4
single sentence 7.616389999999999E-4
training scenarios 7.59485E-4
decisions field 7.57517E-4
record types 7.544100000000001E-4
weather forecast 7.52719E-4
greedy approach 7.5055E-4
multiple record 7.43064E-4
terministic function 7.40007E-4
cal weather 7.384679999999999E-4
jective function 7.3616E-4
single record 7.36116E-4
previous record 7.360789999999999E-4
weather forecasts 7.34801E-4
features 7.33452E-4
other domains 7.3332E-4
semantic 7.3211E-4
local weather 7.31988E-4
possible values 7.31311E-4
common weather 7.28001E-4
field sets 7.25874E-4
robocup results 7.241400000000001E-4
joint task 7.22851E-4
