There is no better feeling than to create that killer app that will boost your company (and your career) to new heights. With VoiceXML, the window of opportunity is open for you to make such a leap.
If you are seriously interested in trying out VoiceXML for a test case, look no further than VoiceGenie and their VoiceGenie Applications. The products are straightforward, easy to use, and create tangible results quickly and easily. I have developed a test case to have this article accessed over the phone using VoiceXML, so follow me as I detail the steps to bring a VoiceXML application to life.
The first thing to do, as with any project, is to plan, plan, plan. Following the guide of a VoiceGenie tutorial, I outlined the steps I would need to take to accomplish my task, which is as follows:
1) Preparation Stage; Gather the required tools to develop the app
2) Design Stage;
3) Implementation Stage;
4) Testing Stage;
The Preparation Stage As with any project, the tools must be available to make the app work. At VoiceGenie’s homepage, you will find all you need to create a real, workable VoiceXML project. By signing up to their developer’s workshop, you can access tutorials and code to help you along the way. However, you definitely need the trial version of the Genie IDE, MyGenie, and the gateway to give your site a phone number and extension. VoiceGenie offers up to twenty extensions, free of charge, to test your apps.
The Genie IDE is a development platform that works just like any other HTML app, with pull-down menus of tags and spreadsheet style boxes for attributes inside tags. It also properly formats the page, as well as has an excellent debugger. Although I’ve only been using it a week, I feel very comfortable with the format. Although the help menu is somewhat lacking, and I still can’t enter any input into the run-time, it is very good for the trail version.
The MyGenie is a Gateway that can validate and run your apps. It does take responses, either using a microphone or the keypad. It does require a server, but the page has links to Apache downloads. The only problem I have is that it consistently gives me an unhandled exception when I close the program. Other than that, it is a good way to test the apps you develop for navigability before you upload it. Also, you will need a speech engine to run your apps. VoiceGenie has a link to download Microsoft Speech SDK 5.0, but be prepared to wait if you have a slow connection.
Another couple of tools that I have found helpful is the complete White Paper on the tags from the VoiceXML Forum, and the examples at the VoiceGenie Workshop. Also, there is a grammar wizard to develop grammars for your apps.
Design Stage Because of scheduling deadlines, I decided to set up my design as a triage to ensure the basic needs were met. I set up the triage as Needs, Wants, and Wishes. The Needs are that the page be able to read the title, sponsor, Editor’s Note, and main part of the article. The Wants is a menu that can navigate the user to any of these four parts, and the Wish is to have a command to repeat a paragraph, or to return you to the menu.
Therefore, it should follow this flow diagram:
<vxmldesign.gif>
Also, throughout any paragraph, the user should be able to utter a command to return to the beginning of the paragraph:
<vxmlpara.gif>
Basically, the following should happen:
1) The user calls in, hears the welcome prompt, and then goes to the menu.
2) User chooses a selection by stating the name, or by using the touch tone keys.
3) The user listens to the information, then returns to the menu.
4) If the user wishes to exit during a reading, he can say “Main” to return to the menu.
5) If he wishes for a paragraph to be repeated, he can say “Repeat” to begin the audio at the beginning of the paragraph.
Construction Stage
To begin, I opened a new file with Genie IDE. I began my opening a <menu> tag, which I set the dtmf to “true”, to allow for dialing, and the scope to “document”, so that I can hide the “Main” command in the code. I open with the welcome audio, and then use a menu audio to prompt the choices. When the user inputs a number or says the corresponding choice, it takes the person to the form containing that audio clip. If the user says “Main”, it returns them to the menu prompt. Other than that, the code is straightforward. I have decided to skip the repeat paragraph option for now, as to save time.
Audio Prompts Because this is the first of its kind, I had to use generic prompts to develop the system, while writing this. It was quite a Paradox. However, when it is finished, someone will read the article, and it will be recorded using Microsoft Recorder that comes with all Windows Suites. There is no need to use a sophisticated recording system, because VoiceXML can only use PCM 8 bit Mono formats, usually with .wav or .vox files.
Testing the app Using the MyGenie Gateway, I tested the app for navigatibility and clarity. I then uploaded the page and audio sources to a server, and established a link from VoiceGenie’s Gateway to the vxml page. Now the real test comes to bear. I know personally how to navigate and use the app, but can you?? To give it a try, call 1-(416)-736-9731, ext. 125682 to hear and use this app to have someone narrate this page to you, and tell us what you think. We are always looking for ways to improve our new technologies here at Murdok.
John Mark Kennedy is currently a Senior at the University of Kentucky pursuing a degree in Descision Sciences and Information Systems. His e commerce interests include eCRM and supply chain management infrastructure.
