A Cache-Based Natural Language Model for Speech Recognition
IEEE Transactions on Pattern Analysis and Machine Intelligence
Computation of term associations by a neural network
SIGIR '93 Proceedings of the 16th annual international ACM SIGIR conference on Research and development in information retrieval
Class-based n-gram models of natural language
Computational Linguistics
Corrections to "A Cache-Based Language Model for Speech Recognition"
IEEE Transactions on Pattern Analysis and Machine Intelligence
Distributional clustering of English words
ACL '93 Proceedings of the 31st annual meeting on Association for Computational Linguistics
Similarity-based estimation of word cooccurrence probabilities
ACL '94 Proceedings of the 32nd annual meeting on Association for Computational Linguistics
Neural network approach to word category prediction for English texts
COLING '90 Proceedings of the 13th conference on Computational linguistics - Volume 3
Automatic model refinement: with an application to tagging
COLING '94 Proceedings of the 15th conference on Computational linguistics - Volume 1
Recent topics in speech recognition research at NTT laboratories
HLT '91 Proceedings of the workshop on Speech and Natural Language
Using word support model to improve Chinese input system
COLING-ACL '06 Proceedings of the COLING/ACL on Main conference poster sessions
Hi-index | 0.00 |
The runtime application domain has a great effect on the performance of practical corpus-based applications. Previous smoothing techniques and class-based and similarity-based models could not handle the dynamic status perfectly. In this paper, an adaptive learning algorithm is proposed for task adaptation that best fits the runtime application domain in applying Chinese homophone disambiguation. The proposed algorithm is first formulated by a neural network model and then generalized to avoid the problem of slow convergence. The resulting techniques are greatly simplified and robust. The experimental results demonstrate the effects of the learning algorithm from a generic domain to a specific one. A methodology is also presented to show how these techniques can be extended to various language models and corpus-based applications.