Search International and National Patent Collections

1. (WO2001037128) A SYSTEM AND ITERATIVE METHOD FOR LEXICON, SEGMENTATION AND LANGUAGE MODEL JOINT OPTIMIZATION

Pub. No.:    WO/2001/037128    International Application No.:    PCT/US2000/041870
Publication Date: Sat May 26 01:59:59 CEST 2001 International Filing Date: Sat Nov 04 00:59:59 CET 2000
IPC: G06F 17/27
G10L 15/18
Applicants: MICROSOFT CORPORATION
Inventors: WANG, Hai-Feng
HUANG, Chang-Ning
LEE, Kai-Fu
DI, Shuo
CAI, Dong-Feng
CHIEN, Lee-Feng
GO, Jianfeng
Title: A SYSTEM AND ITERATIVE METHOD FOR LEXICON, SEGMENTATION AND LANGUAGE MODEL JOINT OPTIMIZATION
Abstract:
A method for optimizing a language model is presented comprising developing an initial language model from a lexicon and segmentation derived from a received corpus using a maximum match technique, and iteratively refining the initial language model by dynamically updating the lexicon and re-segmenting the corpus according to statistical principles until a threshold of predictive capability is achieved.