Abstract

Speech controlled web agents provide a way to build cheap telephony-based information services that can be used to provide various types of^ information available on the Internet. With this graduation project speech technology and web agents have been combined and a generic model for speech controlled web agents has been developed. Furthermore the characteristics of information services that are suited for speech controlled web agents and the potentials and drawbacks of the current available speech technology have been revealed. This has been done by studying the required techniques of speech recognition, spoken language generation and web agents and applying some state-of-the-art tools in these fields into a working prototype.