Skip to main content

Table 3 New training/development queries used in the 2014 evaluation and not in the 2012 evaluation, time length per query (in milliseconds), and number of occurrences per query

From: Comparison of ALBAYZIN query-by-example spoken term detection 2012 and 2014 evaluations

Query (time)–(# occ.) Query (time)–(# occ.)
Presentación (760)–(17) Portugal (340)–(4)
Vosotros (360)–(6) Parlamento (360)–(3)
Etcétera (490)–(28) Microsoft (580)–(4)
Empresas (820)–(71) Mavir (420)–(2)
Porcentaje (490)–(6) Málaga (450)–(2)
Experimentos (400)–(10) Isabel (310)–(4)
Noventa (630)–(39) Garner (320)–(3)
Atención (280)–(8) Galicia (520)–(4)
Mercado (510)–(111) Erasmus (430)–(2)
Resolver (500)–(8) Dilbert (640)–(2)
Probablemente (490)–(6) Complutense (460)–(4)
Dominios (370)–(17) Cristian (510)–(2)
Wikipedia (670)–(3) Berrilan (430)–(2)
Webmaster (460)–(2) Aguilera (480)–(3)
Valladolid (530)–(2) Premios nobel (650)–(2)
Sevilla (450)–(2) Universidad de Chile (840)–(3)
Profit (330)–(3) Nick cohn (410)–(3)