Table 4 Development and test term list characteristics for MAVIR, RTVE, and COREMAH databases
From: ALBAYZIN 2018 spoken term detection evaluation: a multi-domain international evaluation in Spanish
Term list | dev-MAVIR | dev-RTVE | test-MAVIR | test-RTVE | test-COREMAH |
|---|---|---|---|---|---|
#IN-LANG terms (occ.) | 354 (959) | 307 (1151) | 208 (2071) | 301 (1082) | 153 (1022) |
#OUT-LANG terms (occ.) | 20 (55) | 91 (351) | 15 (50) | 103 (162) | 8 (16) |
#SINGLE terms (occ.) | 340 (984) | 380 (1280) | 198 (2093) | 383 (1186) | 145 (1004) |
#MULTI terms (occ.) | 34 (30) | 18 (222) | 25 (28) | 21 (58) | 16 (34) |
#INV terms (occ.) | 292 (668) | 312 (1263) | 192 (1749) | 316 (1035) | 128 (948) |
#OOV terms (occ.) | 82 (346) | 86 (239) | 31 (372) | 88 (209) | 33 (90) |