Modeling Long Distance Dependence in Language: Topic Mixtures vs. Dynamic Cache Models (1996)
| Venue: | IEEE Transactions on Speech and Audio Processing |
| Citations: | 77 - 1 self |
BibTeX
@INPROCEEDINGS{Iyer96modelinglong,
author = {R. Iyer and M. Ostendorf},
title = {Modeling Long Distance Dependence in Language: Topic Mixtures vs. Dynamic Cache Models},
booktitle = {IEEE Transactions on Speech and Audio Processing},
year = {1996},
pages = {236--239}
}
Years of Citing Articles
Bookmark
 |
 |
 |
 |
 |
 |
OpenURL
Abstract
In this paper, we investigate a new statistical language model which captures topic-related dependenciesof words within and across sentences. First, we develop a sentence-level mixture language model that takes advantage of the topic constraints in a sentence or article. Second, we introduce topic-dependent dynamic cache adaptation techniques in the framework of the mixture model. Experiments with the static (or unadapted) mixture model on the 1994 WSJ task indicated a 21% reduction in perplexity and a 3-4% improvement in recognition accuracy over a general n-gram model. The static mixture model also improved recognition performance over an adapted n-gram model. Mixture adaptation techniques contributed a further 14% reduction in perplexity and a small improvement in recognition accuracy.
