Voice XML Starter Pack

 

This information should help you get started with using or developing the Lucidium Voice Interface.

 

Background to the Voice Demo

We would like to create a voice skin to the Lucidium Application Platform. Once delivered voice users could access, search and navigate any relational data model and most CRUD processes without requiring the use of code or significant development effort.

 

Testing the Unfinished Voice Demo

Test demo account using 1 of the following methods.

 

 

Editing the Demo

Control and edit demo account at BeVocal Café here

 

 

There are 2 files loaded

 

 

Do not think edit the VXML and think that the job is done. See below!

 

This is XML and XSL not vXML!

That was an example only. What you tested were STATIC vxml files but that this is not how the system will work.

 

Anybody can create static files. We need to generate a configuration and data driven dynamic interface. We are close.

 

The vXML you saw was generated not written. It was generated by XML/XSL transformation done locally and manually. Once development is done (best done locally I found) then the file will be loaded to the server and this transformation must happen for any page each time that page is called from the voice interface. How and why that happened (serverside XSL transformation and device rules) is out of scope and already delivered.

 

The Process

The following will be the process.

 

  1. An HTTP request is made to the server.
  2. The server creates an XML packet that contains the requested data. The document conforms to the schema in the zip file.
  3. The server notes that this is a voice interface and applies the sy_global.xsl transformation to dynamically generate vXML for delivery
  4. The BeVocal or equivalent platforms renders the vXML

 

The System Files

This document makes reference to a zipped archive containing example files. If folks take an interest in undertaking development work in this area then they and their work will be integrated into the standard Lucidium/SourceForge CVS.

 

Download the example files here

 

 

The zip includes 2 XSL files. The “sy_global html.xsl” file is for reference only and is the XML>HTML equivalent of the file we are trying to create for vXML. It takes any XML generated by the application and renders it as HTML. The other starts to do the same for voiceXML. It is not complete.

 

The XML files are example XML bundles. Each should be able to be viewed both as HTML (see the screenshots) and as vXML. Once that is possible we are nearly done.

 

Other XML

All pages can be accessed either in HTML or XML mode as follows.

 

Example - 1 Project

 

http://lucidium.org.uk:8080/lu_dvt/lu?_ENV=default&_MENU=pj&_UID=guest&_PWD=guest&_ENV=default

 

http://lucidium.org.uk:8080/lu_dvt/lu?_ENV=default&_MENU=pj&_UID=guest&_PWD=guest&_ENV=default&_XML=

 

Local Transformation, BeVocal Testing

 

To see how close we are transform an XSL into vXML using the sy_global.xsl file and paste into the BeVocal testing environment. You will see that although the menu and page outline is in place the content is still lacking.

 

Progress and Contribution

We want your help. We think that this could be interesting. A voice interfave to ANY relational data molde without the need for code and script.

 

We need your help to design flows, dialogs, pages. If you can write XSL then maybe you have a contribution there too.

 

Anyone able to help would quickly become a valued member or the team and will be able to claim to have an understanding of and a hand in the development of a breakthrough technology.

 

Review this technology and email robert.neild@lucidium.net me if you have questions or would like to contribute to this project.