Measuring the acceptable word error rate of machine-generated webcast transcripts
The DOI matches its Crossref record; 19 of the 20 references it cites were found in a registry or an index.
Abstract
The increased availability of broadband connections has recently led to an increase in the use of Internet broadcasting (webcasting). Most webcasts are archived and accessed numerous times retrospectively. One of the hurdles users face when browsing and skimming through archives is the lack of text transcripts of the audio channel of the webcast archive. In this paper, we proposed a procedure for prototyping an Automatic Speech Recognition (ASR) system that generates realistic transcripts of any desired Word Error Rate (WER), thus overcoming the drawbacks of both prototypebased and Wizard of Oz simulations. We used such a system in a study where human subjects perform question-answering tasks using archives of webcast lectures, and showed that their performance and perception of transcript quality is linearly affected by WER, and that transcripts of WER equal or less than 25 % would be acceptable for use in webcast archives.
Overview
Integrity signals
Registry
Cites
Statistics
Verification
Claims the paper makes
Claims other papers make about this one
Each is checked against this paper’s own text: does it say what the citing paper says it does?
In the citing paper
“In ASR, a 5% WER, over longer segments of text, is standard for professional-level transcriptions141 and 25% or less is considered acceptable142.”
From The speech neuroprosthesis (2024) · cites this paper as [142]
Citations
Cites
Wizard of Oz studies: why and how
Dahlbäck, N., Jönsson, A., & Ahrenberg, L. (1993) · Proceedings of the 1st international conference on Intelligent user interfaces - IUI '93 · doi:10.1145/169891.169968 · Cited by 742
Statistical language modeling using the CMU-cambridge toolkit
Clarkson, P., & Rosenfeld, R. (1997) · 5th European Conference on Speech Communication and Technology (Eurospeech 1997) · doi:10.21437/eurospeech.1997-683 · Cited by 563
Nomadic radio: speech and audio interaction for contextual messaging in nomadic environments
Sawhney, N., & Schmandt, C. (2000) · ACM Transactions on Computer-Human Interaction · doi:10.1145/355324.355327 · Cited by 298
SpeechSkimmer: a system for interactively skimming recorded speech
Arons, B. (1997) · ACM Transactions on Computer-Human Interaction · doi:10.1145/244754.244758 · Cited by 182
Designing Interactive Speech Systems: From First Ideas to User Testing
Bernsen, N. O., & Dybkjær, L. (1998) · Medical Entomology and Zoology · Cited by 177
Designing Interactive Speech Systems
Bernsen, N. O., Dybkjær, L., & Dybkjær, H. (1998) · Springer eBooks · doi:10.1007/978-1-4471-0897-9 · Cited by 161
SCANMail: a voicemail interface that makes speech browsable, readable and searchable
Whittaker, S., Hirschberg, J., Amento, B., Stark, L., Bacchiani, M., Isenhour, P., Stead, L., Zamchick, G., & Rosenberg, A. (2002) · Proceedings of the SIGCHI Conference on Human Factors in Computing Systems · doi:10.1145/503376.503426 · Cited by 93
The effect of speech recognition accuracy rates on the usefulness and usability of webcast archives
Munteanu, C., Baecker, R., Penn, G., Toms, E., & James, D. (2006) · Proceedings of the SIGCHI Conference on Human Factors in Computing Systems · doi:10.1145/1124772.1124848 · Cited by 79
Self-citationUser acceptance of handwritten recognition accuracy
LaLomia, M. J. (1994) · CHI Conference Extended Abstracts on Human Factors in Computing Systems (CHI EA) · doi:10.1145/259963.260086 · Cited by 78
Park, A., Hazen, T. J., & Glass, J. R. (2006) · IEEE International Conference on Acoustics Speech and Signal Processing · doi:10.1109/icassp.2005.1415159 · Cited by 76
Cited by
Machine translation of cortical activity to text with an encoder–decoder framework
Makin, J. G., Moses, D. A., & Chang, E. F. (2020) · Nature Neuroscience · doi:10.1038/s41593-020-0608-8 · Cited by 321
“For reference, in speech transcription, WERs of 5% are professional-level 21 , and 20–25% is the outer bound of acceptable performance 22 .”
BackgroundSilva, A. B., Littlejohn, K. T., Liu, J. R., Moses, D. A., & Chang, E. F. (2024) · Nature Reviews Neuroscience · doi:10.1038/s41583-024-00819-9 · Cited by 103
“For text-decoding systems, these metrics draw heavily on precedents set in ASR 141,142 .”
Combining multiple thresholding binarization values to improve OCR output
Lund, W. B., Kennard, D. J., & Ringger, E. K. (2013) · Proceedings of SPIE, the International Society for Optical Engineering/Proceedings of SPIE · doi:10.1117/12.2006228 · Cited by 57
Towards More Robust Speech Interactions for Deaf and Hard of Hearing Users
Fok, R., Kaur, H., Palani, S., Mott, M. E., & Lasecki, W. S. (2018) · Proceedings of the 20th International ACM SIGACCESS Conference on Computers and Accessibility · doi:10.1145/3234695.3236343 · Cited by 34
An embedded end-to-end voice assistant
Lazzaroni, L., Bellotti, F., & Berta, R. (2024) · Engineering Applications of Artificial Intelligence · doi:10.1016/j.engappai.2024.108998 · Cited by 32
Metadata
- Authors
- Munteanu, C., Penn, G., Baecker, R., Toms, E., & James, D.
- Affiliations
- University of Toronto (CA)
- Dalhousie University (CA)
- Venue
- Interspeech 2006
- Publisher
- ISCA
- Published
- 2006-09-17
- Pages
- paper 1756-Mon1CaP.2-0
- ISSN
- 2308-457X (ISSN-L)
- OpenAlex
- W34011602
- Cited by
- 21
- References
- 20
- Citations by year
- 2026: 1 · 2025: 2 · 2024: 4 · 2021: 1 · 2020: 2 · 2018: 1 · 2013: 3 · 2012: 2
- FWCI
- 0.72
- Type
- conference paper
- Language
- en
- Open access
- closed
- Locations
- Interspeech (USB) (publishedVersion)
- Topics
- Physical Sciences › Computer Science › Artificial Intelligence › Speech Recognition and Synthesis
- Physical Sciences › Computer Science › Signal Processing › Speech and Audio Processing
- Physical Sciences › Computer Science › Artificial Intelligence › Natural Language Processing Techniques
- Keywords
- internet broadcasting, word error rate, speech recognition, question answering, user perception, Wizard of Oz simulation
About these checks
Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.
Statistics and arithmetic are recomputed, not judged.
Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.
What counts as a claim ›How we measure confidence ›How we check this ›
Are you an author, or do you see a problem with this page? Request a review
How to cite
Cite CoChat for the checks on this page, and the paper for its own findings.
- This page
- CoChat. (2026). Citation check: Measuring the acceptable word error rate of machine-generated webcast transcripts. https://app.cochat.ai/papers/10.21437/interspeech.2006-40
- The paper
- Munteanu, C., Penn, G., Baecker, R., Toms, E., & James, D. (2006). Measuring the acceptable word error rate of machine-generated webcast transcripts. Interspeech 2006. https://doi.org/10.21437/interspeech.2006-40
Data sources
Bibliographic data: Semantic Scholar, OpenAlex, Crossref.
Contains information from the Semantic Scholar Academic Graph Datasets, made available under the ODC Attribution License (ODC-BY 1.0).
Semantic Scholar’s data is described in: Kinney, R., et al. (2023). The Semantic Scholar Open Data Platform. arXiv:2301.10140.