| J13-2001 | from two given NE lists , the | NE translation | model is usually adopted . Typically |
| N04-1036 | phonetic similarity to improve the | NE translation | accuracy . As shown in the previous |
| N04-1036 | apply these measures to search for | NE translation | pairs using the cross-lingual |
| J13-2001 | abbreviation cases are not rare in | NE translation | . We believe that an expansion |
| J13-2001 | insertions and deletions during | NE translation | . As an example , shown in Table |
| H05-1061 | estimate both and , and define the | NE translation | cost as : ) | ( lj efp ) | ( |
| N04-1036 | document , the query for a target | NE translation | search can be flexible : a few |
| J13-2001 | the NE recognition model and the | NE translation | table . Additionally , the impact |
| N04-1036 | documents may not contain the target | NE translation | . On the other hand , long queries |
| J13-2001 | 2004 ) proposed to extract the | NE translation | dictionary from the bilingual |
| J13-2001 | tagging features ) , from which a | NE translation | dictionary was then constructed |
| I05-2031 | golden standard translations . | NE translation | involves both semantic translation |
| J13-2001 | performance of ruleapplication in | NE translation | was not described , however . |
| N04-1036 | relevant documents and correct | NE translations | is also reduced . Due to the |
| J13-2001 | extract a smaller but cleaner | NE translation | dictionary and then used the |
| J13-2001 | can then be used to train the | NE translation | model ) . Furthermore , with |
| J13-2001 | the first step in producing the | NE translation | table ( which can then be used |
| N04-1036 | measures have been introduced to find | NE translation | pairs : pronunciation similarity |
| N04-1036 | pronunciations , it is possible to discover | NE translation | pairs through their written forms |
| J14-2008 | framework is proposed for extracting | NE translation | pairs from both comparable and |
