Paper: Language Model Adaptation With Map Estimation And The Perceptron Algorithm

ACL ID N04-4006
Title Language Model Adaptation With Map Estimation And The Perceptron Algorithm
Venue Human Language Technologies
Session Short Paper
Year 2004
Authors

In this paper, we contrast two language model adaptation approaches: MAP estimation and the perceptron algorithm. Used in isolation, we show that MAP estimation outperforms the lat- ter approach, for reasons which argue for com- bining the two approaches. When combined, the resulting system provides a 0.7 percent ab- solute reduction in word error rate over MAP estimation alone. In addition, we demonstrate that, in a multi-pass recognition scenario, it is better to use the perceptron algorithm on early pass word lattices, since the improved error rate improves acoustic model adaptation.