Inteligentné rečové komunikačné rozhranie

The system Inteligentné rečové komunikačné rozhranie (IRKR) mediate voice interaction between human and computer. It enables multi-user interaction in Slovak language to find information distributed in Internet or in database.



IRKR is based on Galaxy Communicator infrastructure and HUB architecture. IRKR consist of HUB and five servers, which co-operate together - Audio Server, Automatic speech Recognize Server (ASR), Text to Speech Server (TTS), Dialog Manager (DM) and Information Server (IS).

• Audio Server - supports the hardware part of IRKR and adject to telecommunication network.
  
• ASR Server (Automated Speech Recognition) - performs conversion of incoming speech to corresponding text.
  
• TTS Server (Text to Speech) - synthesis converts outgoing text to speech, which is better suited for user.
  
• Information server - acquire the information required by user from database or internet.
  
• Dialog manager - manage the dialog for a consideration of interpret VoiceXML mark up language from VXML documents.
  
• HUB - central server, which manage associations between servers.

The dialog must be articulate, simple, look like as dialog between 2 persons and consist of several elementary steps. At first Audio server informs the Hub that audio is available, and the Hub informs the Recognizer to start processing. Than ASR sent message to HUB, for configuration of other servers. In next step HUB sent message to DM for start dialog. ASR server recognizes user input and DM consult Backend server and retrieve a database respond. IS find necessary information and set it back. DM decides that it's time to say something to the user. So it sends a new message to the Hub, which is routed to the TTS. Finally, the TTS server produces a new message to notify the Audio server that audio is available, and the Audio server fetches the audio. User can heart required information.