W15-1523 |
one of its synonyms . For the
|
list extraction
|
setup , the training word pairs
|
W10-1102 |
and creatinine . The medication
|
list extraction
|
process uses a manually created
|
N07-2015 |
Related work The idea of n-best
|
list extraction
|
from a word graph for SMT was
|
N07-2015 |
described an efficient n-best
|
list extraction
|
method that is based on the k
|
W05-0834 |
present ef cient pruning and N-best
|
list extraction
|
techniques . In Section 4 , we
|
N07-2015 |
timeconsuming than the actual n-best
|
list extraction
|
. The average generation time
|
W15-1523 |
synonyms is 0.020 , while for the
|
list extractions
|
, it is 0.040 . Both lead to
|
W15-1523 |
although there is some noise , the
|
list extraction
|
also captures semantically meaningful
|
W05-0834 |
full path scores . + The N-best
|
list extraction
|
does not eliminate duplicates
|
W05-0834 |
-LSB- ? -RSB- C / t 3.4 N-Best
|
List Extraction
|
In this section , we describe
|
W10-1744 |
scoring element , until k-best
|
list extraction
|
when it is lazily unpacked .
|
N07-2015 |
the simplex method . 2.2 N-best
|
list extraction
|
We incorporated an efficient
|
W02-1207 |
growth is studied by using Word
|
list Extraction
|
tool to extract word lists from
|
N07-2015 |
2006 ) and enhance the n-best
|
list extraction
|
with Eppstein 's k shortest path
|
W06-3119 |
submission , we modified the K-Best
|
list extraction
|
process to integrate an n-gram
|
J14-4002 |
lattice rescoring rather than n-best
|
list extraction
|
. Similar problems arise when
|
W06-3119 |
language model into the K-Best
|
list extraction
|
process . Our final system shows
|
D10-1115 |
site . We performed some of the
|
list extraction
|
and checking operations we are
|
W15-1523 |
semantic proximity of words . 4.2
|
List Extraction
|
Lists and enumerations are another
|