Combining Slot-based Vector Space Model for Voice Book Search

We describe a hybrid approach to vector space model that improves accuracy in voice search for books. We compare different vector space approaches and demonstrate that the hybrid search model using a weighted sub-space model smoothed with a general model and a back-off scheme provides the best search performance on natural queries obtained from the Web.

[1]  Geoffrey Zweig,et al.  Automated directory assistance system - from theory to practice , 2007, INTERSPEECH.

[2]  Dong Yu,et al.  An introduction to voice search , 2008, IEEE Signal Processing Magazine.

[3]  Alexander I. Rudnicky,et al.  Using the Amazon Mechanical Turk for transcription of spoken language , 2010, 2010 IEEE International Conference on Acoustics, Speech and Signal Processing.

[4]  R. Bashirullah,et al.  Technology and Signal Processing for Brain-Machine Interfaces , 2008, IEEE Signal Processing Magazine.

[5]  Young-In Song,et al.  Voice search of structured media data , 2009, 2009 IEEE International Conference on Acoustics, Speech and Signal Processing.

[6]  Wayne H. Ward,et al.  Recent Improvements in the CMU Spoken Language Understanding System , 1994, HLT.

[7]  Susan L. Epstein,et al.  Learning about Voice Search for Spoken Dialogue Systems , 2010, NAACL.

[8]  Geoffrey Zweig,et al.  Voice-Rate: A Dialog System for Consumer Ratings , 2007, HLT-NAACL.

[9]  Ellen M. Voorhees,et al.  The TREC-8 Question Answering Track Evaluation , 2000, TREC.

[10]  Gary Geunbae Lee,et al.  Data-driven user simulation for automated evaluation of spoken dialog systems , 2009, Comput. Speech Lang..