| W05-0610 | processed using the open-source | ANNIE system | , which is part of GATE ( Cunningham |
| E06-1050 | predefined set of answer types . The | ANNIE system | represents a more realistic expectation |
| E06-1050 | oracle system , our model , and the | ANNIE system | broken down by manually-assigned |
| P99-1040 | appropriate in dialogue D1 from the | Annie system | ( Kamm et al. , 1998 ) , shown |
| N13-1027 | English are processed with the | ANNIE system | , a morphological an - alyzer |
| P02-1022 | , emails etc. ) . The default | ANNIE system | produces figures of between 80-90 |
| W02-0403 | same as those used in the default | ANNIE system | . The difference between Company |
| W03-1505 | baseline system , the default | ANNIE system | for English , which we ran on |
| W03-0803 | available with GATE , as part of the | ANNIE system | ( Cunningham et al. , 2002 ) |
| W03-1505 | data available , we adapted the | ANNIE system | for Cebuano and achieved an Fmeasure |
| W03-0803 | gazetteer lists provided by the | ANNIE system | . Tokens were used as instance |
| E06-1050 | contexts . Our model outperforms the | ANNIE system | in general , and for questions |
| S07-1063 | analysed by the default GATE1 | ANNIE system | ( Cunningham et al. , 2002 ) |
| E03-2013 | purpose components are part of the | ANNIE system | , distributed with GATE : • |
