Utterance classification with discriminative language modeling
Speech Communication, cilt.48, sa.3-4, ss.276-287, 2006 (SCI-Expanded, Scopus)
- Yayın Türü: Makale / Tam Makale
- Cilt numarası: 48 Sayı: 3-4
- Basım Tarihi: 2006
- Doi Numarası: 10.1016/j.specom.2005.06.010
- Dergi Adı: Speech Communication
- Derginin Tarandığı İndeksler: Science Citation Index Expanded (SCI-EXPANDED), Scopus
- Sayfa Sayıları: ss.276-287
- Anahtar Kelimeler: speech recognition, language modeling, discriminative training, utterance classification
- Boğaziçi Üniversitesi Adresli: Evet
Özet
This paper investigates discriminative language modeling in a scenario with two kinds of observed errors: errors in ASR transcription and errors in utterance classification. We train joint language and class models either independently or simultaneously, under various parameter update conditions. On a large vocabulary customer service call-classification application, we show that simultaneous optimization of class, n-gram, and class/n-gram feature weights results in a significant WER reduction over a model using just n-gram features, while additionally significantly outperforming a deployed baseline in classification error rate. A range of parameter estimation approaches, based on either the perceptron algorithm or conditional log-linear models, for various feature sets are presented and evaluated. The resulting models are encoded as weighted finite-state automata, and are used by intersecting the model with word lattices. © 2005 Elsevier B.V. All rights reserved.