Research And Implemenation Of Voice Intelligent Plantform Based On Voicexml

Interactive Voice Response (IVR) is a device which is responsible for the interaction between call centers and clients and providing self-service. Conventional IVR systems are mainly key-pressed and use self-defined script language to realize the complex interaction. This solution has some disadvantages:script languages cannot be unified and cannot be closely combined with Web service, and products lack competiveness. With the development of voice technology, Human-computer Speech Interaction is an inevitable trend in the development of IVR technology. Voice eXtensible Markup Language (VoiceXML) is a Web-based Markup Language which is used to describe the speech between human and computers and describe the voice process.Based on the research of the traditional IVR technique, this thesis adopts the VoiceXML solution, and uses text-to-speech (TTS) technique and automated speech recognition (ASR) technique, in an attempt to study and implement the voice Intelligent Interactive Platform based on the VoiceXML. The voice Intelligent Interactive Platform takes the VoiceXML as the standard script language and interacts with people by using browsers and telephones. It has the TTS and ASR function. Based on the open source project (OpenVXI), the voice Intelligent Interactive Platform has realized its integration with TTS engine and ASR engine, and has completed the parse of the VoiceXML and the design and development of the interface of speech function, which can achieve human-computer speech interaction. Furthermore, the subset of VoiceXML is defined in the Intelligent Interactive Platform, and it has been parsed with the use of MSXML parser so as to realize the navigation function and service function of the voice Intelligent Interactive Platform. The service function can be customized according to the different application requirements and the intelligent service function has been achieved based on the intelligent strategy. Finally, this thesis designs a case study to test the voice intelligent interactive platform.