Automatic allocation of training data for speech understanding based on multiple model combinations

Kazunori Komatani, Mikio Nakano, Masaki Katsumaru, Kotaro Funakoshi, Tetsuya Ogata, Hiroshi G. Okuno

Research output: Contribution to journalArticle

Abstract

The optimal way to build speech understanding modules depends on the amount of training data available. When only a small amount of training data is available, effective allocation of the data is crucial to preventing overfitting of statistical methods. We have developed a method for allocating a limited amount of training data in accordance with the amount available. Our method exploits rule-based methods for when the amount of data is small, which are included in our speech understanding framework based on multiple model combinations, i.e., multiple automatic speech recognition (ASR) modules and multiple language understanding (LU) modules, and then allocates training data preferentially to the modules that dominate the overall performance of speech understanding. Experimental evaluation showed that our allocation method consistently outperforms baseline methods that use a single ASR module and a single LU module while the amount of training data increases.

Original languageEnglish
Pages (from-to)2298-2307
Number of pages10
JournalIEICE Transactions on Information and Systems
VolumeE95-D
Issue number9
DOIs
Publication statusPublished - 2012 Sep
Externally publishedYes

Keywords

  • Language understanding
  • Limited amount of training data
  • Rapid prototyping
  • Spoken dialogue system

ASJC Scopus subject areas

  • Software
  • Hardware and Architecture
  • Computer Vision and Pattern Recognition
  • Electrical and Electronic Engineering
  • Artificial Intelligence

Fingerprint Dive into the research topics of 'Automatic allocation of training data for speech understanding based on multiple model combinations'. Together they form a unique fingerprint.

  • Cite this