Name
Abstract | ||
|---|---|---|
We investigate languagemodel (LM) adaptation in a meeting recognition application, where the LM is adapted based on recognition output from relevant prior meetings and partial manual corrections. Unlike previous work, which has considered either completely unsupervised or supervised adaptation, we investigate a scenario where a human (e.g., a meeting participant) can correct some of the recognition mistakes. We find that recognition accuracy using the adapted LM can be enhanced substantially by partial correction. In particular, if all content words (about half of all recognition errors) are corrected, recognition improves to the same accuracy as if completely error-free (manually created) transcriptions had been used for adaptation. We also compare and combine a variety of adaptation methods, including linear interpolation, unigram marginal adaptation, and a discriminative method based on “positive” and “negative” N-grams. |
Year | DOI | Venue |
|---|---|---|
2009 | 10.1109/ICASSP.2009.4960689 | ICASSP |
Keywords | Field | DocType |
recognition output,recognition accuracy,meeting recognition,language model adaptation,recognition error,user feedback.,unsupervised adaptation,recognition mistake,adaptation method,supervised adaptation,language modeling,partial correction,index terms— speech processing,meeting recognition application,unigram marginal adaptation,meeting participant,exploiting user feedback,error correction,interpolation,speech processing,speech recognition,n grams,language model,feedback,automatic speech recognition,natural language processing,telephony,decoding,linear interpolation,natural languages,acoustics,gain,indexing terms,speech | Speech processing,Transcription (linguistics),Pattern recognition,Computer science,Interpolation,Speech recognition,Artificial intelligence,Natural language processing,Linear interpolation,Decoding methods,Discriminative model,Language model | Conference |
ISSN | Citations | PageRank |
1520-6149 | 4 | 0.47 |
References | Authors | |
10 | 3 | |
Name | Order | Citations | PageRank |
|---|---|---|---|
| Dimitra Vergyri | 1 | 373 | 36.97 |
| Andreas Stolcke | 2 | 6690 | 712.46 |
| Gokhan Tur | 3 | 35 | 2.94 |