word segmentation 0.005000589999999999
dictionary word 0.004501641
single word 0.004418566
current word 0.004381886
word boundary 0.004308035
word train 0.004208526
word boundaries 0.004208455
unique word 0.004193107999999999
difficult word 0.004190124
present word 0.004188007
domain corpus 0.001986779
pos tags 0.001968118
unknown words 0.001958942
frequent words 0.001929793
domain data 0.0019118450000000001
neighboring words 0.001875159
training corpus 0.001857975
dictionary feature 0.001799051
training data 0.001783041
pos sequence 0.001775342
same feature 0.001756254
dictionary features 0.001718931
feature set 0.001712846
lexical features 0.001711652
pos tagging 0.001701302
same features 0.001676134
total pos 0.0016036
context information 0.001603402
pos estimation 0.001602867
type feature 0.001597492
pos transition 0.001587411
joint model 0.00157633
words 0.00157628
maximal pos 0.001546874
character sequence 0.00151323
rich feature 0.001509693
feature strings 0.0015038540000000002
feature djk 0.0014889920000000002
transition information 0.0014488510000000001
arbitrary features 0.001445839
target domain 0.001429107
domain adaptation 0.001424295
segmentation result 0.00142286
valuable information 0.001401302
full corpus 0.001394475
structure information 0.0013869800000000001
general domain 0.001371171
character types 0.001370934
segmentation lattices 0.00136706
other methods 0.001366081
tag pair 0.001353838
main training 0.001339411
training time 0.001332968
annotated data 0.001330793
balanced corpus 0.001317248
corpus sizes 0.001311183
learning approaches 0.0013111309999999999
japanese text 0.001298089
strings character 0.001285002
unannotated data 0.001261957
notated data 0.001261742
domain adaptability 0.001235047
learning situation 0.001233847
active learning 0.00123304
domain adapta 0.001231823
realistic domain 0.001230954
feature 0.00120592
web text 0.001198018
cal text 0.00116253
unsegmented text 0.001152189
text sec 0.001143818
annotation methods 0.001141642
other approaches 0.001141413
other tasks 0.001131609
features 0.0011258
accuracy gains 0.001119494
boundary tags 0.001118463
information 0.00111062
measuring accuracy 0.001107698
slow training 0.001102006
new probability 0.001097294
segmentation 0.00109208
model 0.00106719
test joint 0.001061295
pointwise approach 0.001051359
previous methods 0.0010507250000000002
partial annotation 0.001050674
pointwise method 0.001041817
corpus 0.00103107
analysis nlp 0.001027341
neighboring tags 0.001017817
japanese nlp 9.97831E-4
structured approach 9.962460000000001E-4
prediction methods 9.917860000000001E-4
character 9.87068E-4
tion methods 9.80074E-4
markov models 9.70962E-4
previous work 9.56087E-4
domain 9.55709E-4
new domains 9.539119999999999E-4
