Sixth European Conference on Speech Communication and Technology
This paper describes a new kind of language models composed of several local models and a general model linking the local models together. Local models describe more finely subparts of the textual data than a conventional n-gram trained on the complete corpus. They are built on lexical and syntactic criteria. Both local and global models are integrated in a single hidden Markov model. Experiments showed a 14% decrease in perplexity compared to a bigram model on a small corpus of telephonic communications.
Full Paper (PDF) Gnu-Zipped Postscript
Bibliographic reference. Nasr, Alexis / Estéve, Yannick / Béchet, Frédéric / Spriet, Thierry / Mori, Renato de (1999): "A language model combining n-grams and stochastic finite state automata", In EUROSPEECH'99, 2175-2178.