| C04-1005 | method , achieving a relative error | rate reduction | of 8.8 % . In order to provide |
| C04-1005 | also achieves a relative error | rate reduction | of 30 % . In addition , our method |
| C04-1005 | achieves a 37 % relative error | rate reduction | over the " IBM Refined " method |
| C04-1005 | obtained from Table 2 . The error | rate reductions | in Table 3 are omitted . Our |
| C04-1005 | reduction . 6 Discussion The error | rate reductions | in this paragraph are obtained |
| D14-1096 | the state-of-the-art , an error | rate reduction | of 22.3 % . This is despite using |
| C04-1005 | resulting in a 12 % relative error | rate reduction | . 6 Discussion The error rate |
| C04-1076 | 79.38 . This is , 2.04 % error | rate reduction | compared with the 3-state HMM |
| C04-1103 | 20 percent absolute word error | rate reduction | . Thus , although the experiment |
| D09-1135 | reached 87.5 % , with an error | rate reduction | of 10.1 % , relative to the baseline |
| A94-1030 | drop with k = 8 , with an error | rate reduction | of 49.4 % at k = 2 . Thus any |
| D08-1087 | to a 1.2 % absolute word error | rate reduction | over a linearly interpolated |
| C02-1140 | system , representing an error | rate reduction | of 35.8 % . However , the average |
| D11-1103 | best re-scoring , and word error | rate reduction | of 8 % ( relative ) on a highly |
| D14-1219 | the corresponding relative error | rate reduction | ( ERR ) with respect to the baseline |
| E03-1005 | results show a 9 % relative error | rate reduction | . While SL-DOP and LS-DOP have |
| C04-1005 | achieving a 47 % relative error | rate reduction | . Our method also achieves a |
| D15-1275 | corpus tested , the relative error | rate reduction | was 85.85 % . The system including |
| C04-1005 | approach achieves a relative error | rate reduction | of 26 % and 25 % when compared |
| D12-1054 | musical classifiers , with error | rate reductions | of up to 31 % . 1 Introduction |
