word alignment 0.00768866
alignment alignment 0.00627742
word segmentation 0.005846769999999999
hypothesis word 0.00566314
word error 0.005359345
word sequence 0.005301943999999999
reference word 0.005254862999999999
first word 0.005245245
recognizer word 0.005033396999999999
word errors 0.005029214
word recognition 0.004983161
last word 0.004969993
corresponding word 0.004955849
word timestamps 0.00494774
accurate word 0.0049218529999999995
metric word 0.004911934
word correctness 0.004897052
word gingrich 0.004885189999999999
alignment algorithm 0.00440312
different alignment 0.003906572
alignment information 0.0037113569999999998
alignment algorithms 0.003657157
standard alignment 0.0036531439999999997
alignment errors 0.003617974
alignment phase 0.003608621
phonetic alignment 0.0035347359999999997
top alignment 0.003532554
explicit alignment 0.003529806
actual alignment 0.003525682
sclite alignment 0.003525005
tolerance alignment 0.0035078569999999996
alignment process 0.00349431
lexeme alignment 0.003483388
forced alignment 0.003473408
intuituive alignment 0.003473408
alignment possibilities 0.003473408
alignment mistakes 0.003473408
netic alignment 0.003473408
alignment 0.00313871
hypothesis words 0.00257192
test data 0.002139939
first data 0.002102945
speech data 0.0020936540000000003
recognition system 0.0020338310000000003
algorithm scores 0.001996466
system alembic 0.0019434320000000001
alembic system 0.0019434320000000001
bbn system 0.00194012
final words 0.001918343
new algorithm 0.0019028320000000001
content words 0.001844775
identical words 0.001828453
test corpus 0.001817829
hypothesis text 0.001804496
novel words 0.00180345
newswire data 0.0018002480000000002
timestamped words 0.001796201
noisy data 0.0017824440000000002
muc algorithm 0.0017796490000000001
annotated data 0.0017633640000000001
mapping algorithm 0.0017543980000000001
data consortium 0.001745056
guistic data 0.00174386
phonetic algorithm 0.0016604360000000002
sclite algorithm 0.0016507050000000001
news corpus 0.001638974
coarse segmentation 0.001636488
sophisticated algorithm 0.0016080220000000002
system 0.00160062
algorithm proceeds 0.0016003430000000002
hypothesis texts 0.0015521060000000001
hypothesis nes 0.0015322439999999999
hypothesis type 0.001487728
test set 0.001486589
spurious hypothesis 0.001479325
hypothesis element 0.001478001
hypothesis phrases 0.001476469
final score 0.001464033
words 0.00145873
hypothesis documents 0.001455817
hypothesis files 0.00145545
unmapped hypothesis 0.001453319
hypothesis transcriptions 0.001448228
same speech 0.001396821
reference text 0.0013962190000000002
score components 0.001389048
score component 0.0013775670000000001
partial score 0.001345001
annotated training 0.0013124599999999999
segmentation 0.00129682
language processing 0.001295037
small test 0.001293722
recognizer output 0.001284089
natural language 0.001269113
recall error 0.0012645109999999998
algorithm 0.00126441
single reference 0.0012604410000000002
error evaluation 0.001253655
average sequence 0.0012297459999999999
standard reference 0.001219347
