Automatic Human Utility Evaluation of ASR Systems: Does WER Really Predict Performance?

Benoit Favre (1, 2) , Kyla Cheung (3) , Siavash Kazemian (4) , Adam Lee (5, 6) , Yang Liu (7) , Cosmin Munteanu (4) , Ani Nenkova (8) , Dennis Ochei (8) , Gerald Penn (4) , Stephen Tratz (9) , Clare Voss (9) , Frauke Zeller (10)

1 LIF - Laboratoire d'informatique Fondamentale de Marseille - UMR 6166
2 TALEP - Traitement Automatique du Langage Ecrit et Parlé
3 Columbia University [New York]
4 University of Toronto
5 Department of Chemistry [York, UK]
6 CUNY - City University of New York [New York]
7 University of Texas at Dallas - Department of Computer Science [Dallas]
8 University of Pennsylvania
9 ARL - US Army Research Laboratory-CIS Directorate
10 UCL - University College of London [London]

no abstract

Domains: Computer Science [cs]; Computation and Language [cs.CL]

Benoit Favre, Kyla Cheung, Siavash Kazemian, Adam Lee, Yang Liu, et al.. Automatic Human Utility Evaluation of ASR Systems: Does WER Really Predict Performance?. Interspeech, Lyon (France), 2013, Unknown, Unknown Region. ⟨hal-01194248⟩