| J03-1005 | which takes into account only a | single reference translation | . • PER ( position-independent |
| E12-3001 | be the availability of only a | single reference translation | . This affects the development |
| J03-1005 | English-to-German translation direction , a | single reference translation | for each test sentence is used |
| D15-1148 | carried over a dataset containing a | single reference translation | for each Chinese sentence . In |
| E12-3001 | coreference , the availability of only a | single reference translation | is a real problem . Improving |
| P02-1040 | of times a word occurs in any | single reference translation | . Next , one clips the total |
| J10-3003 | data sets usually have only a | single reference translation | . Madnani et al. ( 2007 , 2008b |
| D09-1106 | sets contain 2,000 sentences with | single reference translations | . We tokenized and lowercased |
| J10-3003 | However , it is impossible for a | single reference translation | to capture all possible verbalizations |
| J14-2005 | the 1,000 sentences , we chose a | single reference translation | from among the four references |
| J11-4002 | stands for the case when only a | single reference translation | is available , and NR > 1 |
| E12-1045 | training . Dev and test sets have a | single reference translation | . For both language pairs , we |
| E12-3001 | simply score the output against a | single reference translation | . To obtain these figures , the |
| P02-1040 | may use a big test corpus with a | single reference translation | , provided that the translations |
| D08-1066 | lations , when compared against a | single reference translation | per sentence . Results : We compare |
| D09-1040 | , and the Chinese source as a | single reference translation | . We tested the system using |
| N06-1002 | held-out sentences each with a | single reference translation | . Parameters were individually |
| E12-3001 | determine whether deviations from the | single reference translation | provided in the PCEDT 2.0 corpus |
| J09-1002 | source sentence is compared with a | single reference translation | and the longest common character |
| E12-1025 | source sentence is compared with a | single reference translation | and the longest common character |
