Hub4 language modeling using domain interpolation and data clustering

Hub4 language modeling using domain interpolation and data clustering (1997)

Cached

Download Links

by Fuliang Weng , Andreas Stolcke , Ananth Sankar

Venue:	In Proceedings DARPA Speech Recognition Workshop
Citations:	5 - 2 self

Versions

Version History

Metadata Version 0

Datum	Value	Source
TITLE	Hub4 Language Modeling Using Domain Interpolation and Data Clustering	SVM HeaderParse 0.2
AUTHOR NAME	Fuliang Weng	SVM HeaderParse 0.2
AUTHOR AFFIL	Speech Technology And Research Laboratory; SRI International	SVM HeaderParse 0.2
AUTHOR ADDR	Menlo Park, California	SVM HeaderParse 0.2
AUTHOR NAME	Andreas Stolcke	SVM HeaderParse 0.2
AUTHOR AFFIL	Speech Technology And Research Laboratory; SRI International	SVM HeaderParse 0.2
AUTHOR ADDR	Menlo Park, California	SVM HeaderParse 0.2
AUTHOR NAME	Ananth Sankar	SVM HeaderParse 0.2
AUTHOR AFFIL	Speech Technology And Research Laboratory; SRI International	SVM HeaderParse 0.2
AUTHOR ADDR	Menlo Park, California	SVM HeaderParse 0.2
ABSTRACT	In SRI’s language modeling experiments for the Hub4 domain, three basic approaches were pursued: interpolating multiple models estimated from Hub4 and non-Hub4 training data, adapting the language model (LM) to the focus conditions, and adapting the LM to different topic types. In the first approach, we built separate LMs for the closely transcribed Hub4 material (acoustic training transcripts) and the loosely transcribed Hub4 material (LM training data), as well as the North-American Business News (NABN) and Switchboard training data, projected onto the Hub4 vocabulary. By interpolating the probabilities obtained from these models, we obtained a 20 % reduction in perplexity and a 1.8 % reduction in word error rate, compared to a baseline Hub4-only language model. Two adaptation approaches are also described: adapting language models to the speech styles correlated with different focus conditions, and building cluster-specific LM mixtures. These two approaches give some reduction in perplexity, but no significant reduction in word error. Finally, we identify the problems and future directions of our work. 1.	SVM HeaderParse 0.2
CITATIONS	4 found	ParsCit 1.0