english words 0.002725441
many words 0.002496348
english word 0.002322921
few words 0.002319037
dutch words 0.002296124
common words 0.002283194
future words 0.002269321
frequent words 0.0022668790000000003
unknown words 0.002234789
unique words 0.002194216
exception words 0.002159752
compound words 0.0021590430000000002
hyphenated words 0.002157207
glish words 0.002153693
incomplete words 0.002153693
words 0.00193895
input word 0.001872872
word forms 0.0018217490000000001
other data 0.001755647
microsoft word 0.0017518660000000001
word appearances 0.0017518660000000001
new method 0.0017348239999999998
english language 0.0016723509999999999
crf method 0.001672217
training data 0.0016172689999999997
different learning 0.001575071
method patterns 0.0015005139999999999
tex method 0.001500432
hyphenation method 0.0014964969999999998
different methods 0.001482328
speed method 0.001401168
talo method 0.0013791489999999999
different values 0.001376511
new english 0.001369815
training set 0.001324537
input data 0.0013202519999999998
learning algorithm 0.001304546
different systems 0.0012861959999999999
different hyphenations 0.00126575
same crf 0.001250336
appropriate data 0.001244519
dutch language 0.001243034
data structure 0.0012270999999999998
language pattern 0.001211957
different versions 0.001206731
different meanings 0.001205075
different lengths 0.001201911
data items 0.0012004379999999999
language processing 0.001199656
data structures 0.0011992919999999998
different pronuncia 0.0011975529999999999
different pronunciations 0.0011975529999999999
learning approach 0.00118743
natural language 0.001170348
crf training 0.0011541759999999998
method 0.0011515
appropriate language 0.001146569
same way 0.001138434
english patterns 0.001135505
lexical information 0.00112167
crf approach 0.001114892
english dataset 0.0011145629999999998
frequent english 0.00111442
large set 0.001110431
american english 0.0010787659999999999
english patgen 0.001073625
tex algorithm 0.001060223
hyphenation algorithm 0.0010562879999999998
entire english 0.001039893
whole english 0.001035483
viterbi algorithm 0.001026325
learning task 0.001024687
crf methods 0.001021229
much information 0.001018423
such function 0.001016807
british english 0.001006578
new text 0.0010047889999999999
same ones 0.0010039860000000001
other researchers 9.92627E-4
high probability 9.91179E-4
other media 9.87208E-4
machine learning 9.64937E-4
same procedure 9.56831E-4
corresponding training 9.50497E-4
same fold 9.47302E-4
development set 9.45493E-4
high threshold 9.42677E-4
tex learning 9.421869999999999E-4
training sets 9.39152E-4
ation algorithm 9.2976E-4
many machine 9.2908E-4
based approach 9.26483E-4
many languages 9.09998E-4
tuning set 9.0869E-4
test error 9.082820000000001E-4
standard crf 9.07922E-4
training list 9.07646E-4
learning problem 9.05806E-4
probability threshold 9.028899999999999E-4
national corpus 9.01424E-4
