A Project of Speech Input and Output in an E-commerce Application

Abstract

The present paper describes the work done during the last year in the development of European Portuguese (EP) speech recognition and synthesis channels for an Internet e-commerce application. The objective of this work was to develop the appropriate speech input and output, to enable the user to more comfortably use an e-commerce web site issuing commands and options under guidance of a menu system. The speech interface operation guidelines are briefly described in their principles and resulting specifications for the speech channels, namely, menu structure and operation dynamics, and in the following, the system software approach and the speech recognition and synthesis modules are presented. A discussion is done about some project trade-offs and results obtained so far. Future perspectives of the on-going work are also presented.KeywordsSpeech RecognitionSpeech InputSynthetic SpeechSegmental QualityEuropean PortugueseThese keywords were added by machine and not by the authors. This process is experimental and the keywords may be updated as the learning algorithm improves.