A dynamic language model for speech recognition
HLT '91 Proceedings of the workshop on Speech and Natural Language
Improvements in stochastic language modeling
HLT '91 Proceedings of the workshop on Speech and Natural Language
Adaptive language modeling using minimum discriminant estimation
ICASSP'92 Proceedings of the 1992 IEEE international conference on Acoustics, speech and signal processing - Volume 1
The Latent Maximum Entropy Principle
ACM Transactions on Knowledge Discovery from Data (TKDD)
Information Retrieval on the Blogosphere
Foundations and Trends in Information Retrieval
A scalable distributed syntactic, semantic, and lexical language model
Computational Linguistics
On the dynamic adaptation of language models based on dialogue information
Expert Systems with Applications: An International Journal
Leveraging relevance cues for language modeling in speech recognition
Information Processing and Management: an International Journal
| Hi-index | 0.00 |
We describe our ongoing efforts at adaptive statistical language modeling. To extract information from the document history, we use trigger pairs as the basic information bearing elements. To combine statistical evidence from multiple triggers, we use the principle of Maximum Entropy (ME). To combine the trigger-based model with the static model, we absorb the latter into the ME formalism. Given consistent statistical evidence, a unique ME solution is guaranteed to exist, and an iterative algorithm exists which is guaranteed to converge to it. Among the advantages of this approach are its simplicity, its generality, and its incremental nature. Among its disadvantages are its computational requirements. We report our current results and discuss possible improvements.