N09-2037 accomplished strictly by means of word deletion . Furthermore , each sequence
D12-1022 single rewrite operation , namely word deletion . Although it is well-known that
H92-1085 were fairly short . One or two word deletions accounted for 82 % of the data
N10-1066 extract features accordingly . For word deletion , we use only the POS tags of
P09-1094 which is mostly reviewed as a word deletion task . Different from prior research
P07-1020 parameter 0 and an additional word deletion penalty can then be used to adjust
P06-1048 2004 ) . Their method is based on word deletion ; given a prespecified compression
J99-4003 recognition that models simple word deletion and repetition patterns . They
N10-1018 were marked as word replacement , word deletion , and word insertion . For details
E99-1003 Gray , 1983 ) rely on function word deletion , stemming , and alphabetical
P06-2019 generates compressions through word deletion . The model does not utilise
H93-1071 satellite . Discourse phrase and word deletion : the signing of a contract or
J10-3008 long-distance movement , and constrained word deletion . We find that these techniques
J15-2001 ( X ) helps us learn a source word deletion model . It is used during decoding
J02-4003 markup errors and no accidental word deletions occurred . We provide a brief
D08-1077 three models for handling " single word deletion " ( they discuss , but do not
H92-1075 errors and consisted of a single word deletion , insertion , or substitution
J10-3005 task is commonly defined as a word deletion problem : Given an input sentence
E14-1022 edit-dist ( M , C ) is the number of word deletions , insertions , and substitutions
H89-2039 correct , word substitution , word deletion , word insertion , and word accuracy
hide detail