Structured Language Modeling for Speech Recognition

Thumbnail Image

Date

Authors

Journal Title

Journal ISSN

Volume Title

Publisher

Abstract

Description

A new language model for speech recognition is presented. The model develops hidden hierarchical syntactic-like structure incrementally and uses it to extract meaningful information from the word history, thus complementing the locality of currently used trigram models. The structured language model (SLM) and its performance in a two-pass speech recognizer --- lattice decoding --- are presented. Experiments on the WSJ corpus show an improvement in both perplexity (PPL) and word error rate (WER) over conventional trigram models.
4 pages + 2 pages of ERRATA

Keywords

Computation and Language, G.3, I.2.7, I.5.1, I.5.4

Citation

URI

http://salesiana.dossiersoluciones.com/handle/123456789/30143

Consulte el texto completo en el siguiente enlace:

https://arxiv.org/abs/cs/0001023
 http://arxiv.org/abs/cs/0001023
 Proceedings of NLDB'99, Klagenfurt, Austria

Collections

arXiv