2000/01/24 by Ciprian Chelba, Frederick Jelinek
Computer Science · #cs.CL
published as Proceedings of Eurospeech, 1999, pp. 1567-1570, Budapest, Hungary · 4 pages
arxiv created 2000/01/24 · arxiv updated 2009/11/30
A new language model for speech recognition inspired by linguistic analysis is presented. The model develops hidden hierarchical structure incrementally and uses it to extract meaningful information from the word history - thus enabling the use of extended distance dependencies - in an attempt to complement the locality of currently used trigram models. The structured language model, its probabilistic parameterization and performance in a two-pass speech recognizer are presented. Experiments on the SWITCHBOARD corpus show an improvement in both perplexity and word error rate over conventional trigram models.