DocumentCode :
3632585
Title :
Bag-of-Words Modelling for Speech Recognition
Author :
Bartosz Ziolko;Suresh Manandhar;Richard C. Wilson
Author_Institution :
Dept. of Electron., AGH Univ. of Sci. & Technol., Krakow, Poland
fYear :
2009
Firstpage :
646
Lastpage :
650
Abstract :
A semantic language modelling method for speech recognition is presented. The method is somehow similar to latent semantic analysis, but it does not need so much memory and training data. Even though it gave better experimental results, provided as percentage of correctly recognized sentences from a corpus. It differentiate from latent semantic analysis by a choice of similar topics influencing a matrix describing probability of words appearing in topics.
Keywords :
"Speech recognition","Natural languages","Sparse matrices","Entropy","Automatic speech recognition","Computer science","Training data","Context modeling","Robustness","Testing"
Publisher :
ieee
Conference_Titel :
Future Computer and Communication, 2009. ICFCC 2009. International Conference on
Print_ISBN :
978-0-7695-3591-3
Type :
conf
DOI :
10.1109/ICFCC.2009.37
Filename :
5189863
Link To Document :
بازگشت