WebSphere Voice Server

WebSphere Voice Server V5.1.x is a software product that can be integrated with other software and hardware telephony products to speech-enable these products and their associated applications.

By using an application developed with WebSphere Voice Server, you can place a telephone call from either a landline or mobile phone.

• A separate telephony server answers this telephone call.
• The telephony server fetches a VoiceXML document from an application server, such as WebSphere Application Server.
• The telephony server then interprets this document and requests either synthesis or recognition to occur on WebSphere Voice Server.
• With WebSphere Voice Server V4.2 for AIX, WebSphere Voice Response acts as the telephony server and communicates with WebSphere Voice Server, which listens to the speaker and recognizes words.
• With WebSphere Voice Server V5.1.x for MRCP, the third-party IVRs and VoiceXML gateway communicate with WebSphere Voice Server, which listens to the speaker and recognizes words.
• These recognized words are then passed to the application, running on an application server such as WebSphere Application Server.
• WebSphere Voice Server then uses the text to synthesize speech, which the telephony server routes back to the handset where the caller can hear it.

The following are examples of how WebSphere Voice Server V5.1.x can be used:

Voice-enabling server applications

• Using open voice standards, software server applications, built on open industry standards, can be enabled for voice access. This allows users to access voice-enabled Web applications by using a telephone either wired or wirelessly.
• You can enable voice on existing DTMF and pre-recorded audio IVR applications with speech recognition and text-to-speech (TTS) capabilities.

Voice-enabling Web applications
You can enable voice on Web applications with a combination of products that includes:

• IBM speech recognition and TTS engines for accepting voice input and generating synthesized speech output
• WebSphere Voice Server application for testing your installation and configuration
• WebSphere Voice Toolkit for WebSphere Studio, which includes a Windows® runtime simulation for Web-based application development

Support software for integration with the following telephony platforms:

• WebSphere Voice Response for AIX, V3.1 or V4.2 (using the included WebSphere Voice Server V4.2 product)
• WebSphere Voice Response V4.2 using the Version 5.1.3 product
• Third-party IVRs and VoiceXML gateways that are qualified and interoperate with the WebSphere Voice Server V5.1.x using MRCP

Speech-enabling IVR application
The speech-enabling IVR application includes:

• IBM speech recognition and TTS engines for recognizing voice input and generating synthesized speech output
• An application for testing your installation and configuration
• Support for system management functions
• Support software for integration with WebSphere Voice Response for AIX V3.1 and V4.2 (WebSphere Voice Server V4.2 only)
• Support software for integration with WebSphere Voice Response V4.2 (WebSphere Voice Server 5.1.3 only)

WebSphere Voice Server V5.1.x product features

• Ability to utilize the J2EE architecture and Enterprise JavaBeans for core functionality. WVS runs as an Enterprise Application in WebSphere Application Server V5.1, extending WebSphere Application Server reliability, scalability, and availability to the WVS 5.1.x server.
• Support for MRCP V1 Draft 4, Speech Recognition Grammar Specification (SRGS) 1.0, and Speech Synthesis Markup Language (SSML) 1.0.
• Support for Semantic Interpretation for Speech Recognition (SISR) - W3C Working Draft 1, dated April 2003.
• Capability to barge-in, which allows a user to interrupt the dialog and respond to a prompt.
• Grammar-based speech recognition, including support for dynamic grammars.
• System administration leveraging WebSphere Application Server Network Deployment V5.1, which includes the WebSphere Application Server System Administrator's Console for ease of configuration, administration, troubleshooting, and reviewing log and trace information. WVS V5.1.x provides additional voice-specific administration panels to facilitate configuration, monitoring, and troubleshooting.
• WebSphere Application Server Network Dispatcher Edge Components (included with WebSphere Application Server Network Deployment) in front of a multi-machine WVS install to allow IP spraying. This allows machines to be taken out of service for maintenance and automatic failover and provides for scaling of the MRCP server.
• WebSphere Application Server Network Deployment Manager, which allows a single WebSphere Application Server System Administrator's Console to manage a network of WVS V5.1.x machines.

WVS V4.2 product features

• Support for VoiceXML 2.0 and VoiceXML 2.1, Speech Recognition Grammar Specification (SRGS) 1.0, and Speech Synthesis Markup Language (SSML) 1.0
• Support for Semantic Interpretation for Speech Recognition (SISR) - W3C Working Draft 1, dated April 2003
• Barge-in, which allows a user to interrupt the dialog and respond to a prompt
• Grammar-based speech recognition, including support for dynamic grammars

Language support
The following information identifies language ASR (Automatic Speech Recognition) and TTS/CTTS (Concatenative TTS) support for each connection:

Note: Note: (WebSphere Voice Server V4.2- WebSphere Voice Response/AIX ):

• U.S. English supports one male and two female voices in third-generation CTTS
• U.K. English, Canadian French, French, German and Spanish support male and female voices in third-generation CTTS
• Japanese supports a female voice in third-generation CTTS
• Korean supports a male voice third-generation CTTS and a ScanSoft female voice
• Dutch, Cantonese, and Italian support ScanSoft TTS (female) only (no formant or concatenative IBM TTS)
• Simplified Chinese supports a CTTS female voice
• Partial SSML support is available (see the VoiceXML Programmer's Guide for specific details)

Note: (WebSphere Voice Server V5.1 - Linux, Windows, and AIX):

• U.S. English supports one male and two female voices in third-generation CTTS and two female voices in fifth-generation CTTS
• U.K. English supports one male and two female voices in third-generation CTTS and 1 female voice in fifth generation CTTS
• Canadian French, German, Latin American Spanish, support male and female voices in third-generation CTTS
• Australian English uses U.K. English CTTS for TTS support
• Simplified Chinese supports a third generation CTTS female voice

Complimentary copies of the following products are included:

• WebSphere Application Server Network Deployment V5.1.1
  This is a restricted license for use by the WVS V5.11x product only
• WebSphere Voice Toolkit V6.0.1 that includes the functionality in Voice Toolkit V5.0 (graphical call flow generation, call control support using CCXML, VoiceXML 2.0 or 2.1 editor, grammar editor, a pronunciation builder, VoiceXML 1.0 to 2.0 conversion utilities, grammar conversion utilities, and an integrated simulator for application testing and debug). These components allow application developers to easily add voice technology to middleware applications.
  
  Note: Although WebSphere Voice Toolkit V6.0.1 supports WVS V5.1.x, WVS V4.2, WVR V4.2, and WVAA V5.0, not all of the preceding capabilities are applicable to each of these products. Additional capabilities provided for WebSphere Voice Server 5.1 support include:
  • Grammar test tool that interfaces with a remote MRCP server
  • Pronunciation migration tools for the new Lexicon file format
  • Lexicon file support for custom pronunciations
• WebSphere Application Server which can be used for internal evaluation and for the development, demonstration, and testing of applications. It provides open standards, the core element in many e-business solution offerings, by offering a rich set of application deployment services and transaction management. A trial version of WebSphere Application Server can be obtained here.
  
  Note: This WebSphere Application Server is not to be used on the same system that WebSphere Voice Server runs on because Voice Server uses the WAS-ND version as included.

Accessibility
WebSphere Voice Server for Multiplatforms is a voice application enabler, which appeals to the major user groups impacted by speech technology-system administrators, application developers, and application end users. WebSphere Voice Server V5.1.x provides the following accessibility features for each user group:

System administrators

• Provides accessible administration and monitoring interfaces for speech technology resources
• Offers text through standard system function calls or through an application programming interface (API), which supports interaction with assistive technology
• Provides accessible documentation

Telephone callers (end users)

• Offers the capability of providing people with visual, manual, and/or mobility impairments access to business data and services using auditory user interfaces
• Helps businesses comply with Section 508 of the Americans with Disabilities Act

Application developers

• Provides an accessible development environment for speech applications.

For technical details, click on the link below:
System requirements