000083156 001__ 83156
000083156 005__ 20180317093239.0
000083156 037__ $$aREP_WORK
000083156 245__ $$aOn the Use of Information Retrieval Measures for Speech Recognition Evaluation
000083156 269__ $$a2004
000083156 260__ $$aMartigny, Switzerland$$bIDIAP$$c2004
000083156 336__ $$aReports
000083156 520__ $$aThis paper discusses the evaluation of automatic speech recognition (ASR) systems developed for practical applications, suggesting a set of criteria for application-oriented performance measures. The commonly used word error rate (WER), which poses ASR evaluation as a string editing process, is shown to have a number of limitations with respect to these criteria, motivating alternative or additional measures. This paper suggests that posing speech recognition evaluation as an information retrieval problem, where each word is one unit of information, offers a flexible framework for application-oriented performance analysis based on the concepts of recall and precision.
000083156 6531_ $$aspeech
000083156 6531_ $$avision
000083156 6531_ $$amccowan
000083156 6531_ $$amoore
000083156 6531_ $$adines
000083156 6531_ $$agatica
000083156 6531_ $$aflynn
000083156 6531_ $$awellner
000083156 6531_ $$abourlard
000083156 700__ $$aMcCowan, Iain A.
000083156 700__ $$aMoore, Darren
000083156 700__ $$0243992$$aDines, John$$g192380
000083156 700__ $$0241066$$aGatica-Perez, Daniel$$g171600
000083156 700__ $$aFlynn, Mike
000083156 700__ $$aWellner, Pierre
000083156 700__ $$0243348$$aBourlard, Hervé$$g117014
000083156 8564_ $$uhttp://publications.idiap.ch/downloads/reports/2004/rr04-73.pdf$$zURL
000083156 8564_ $$s161050$$uhttps://infoscience.epfl.ch/record/83156/files/rr04-73.pdf$$zn/a
000083156 909CO $$ooai:infoscience.tind.io:83156$$preport$$pSTI
000083156 909C0 $$0252189$$pLIDIAP$$xU10381
000083156 937__ $$aEPFL-REPORT-83156
000083156 970__ $$amccowan-rr-04-73/LIDIAP
000083156 973__ $$aEPFL$$sPUBLISHED
000083156 980__ $$aREPORT