We need to identify the cost and effort to integrate voice-to-text and
text-to-voice in front of the application being used at our site. Ex:
A customer should be able to call from his cell phone and get voice 
answers. Answers will be given by a BOT in the form of text. These
answers
will need to be converted into voice through third party tools. The 
answer should also evaluate and list the third party tools available
to
do this. It should state what tools you recommend and why.  It should
also give an idea of any extra hardware needed for this purpose.

---

Request for Question Clarification by wengland-ga on 31 Jul 2002 08:17 PDT

Are you wanting to set up something like SprintPCS has - they have a
manage section on their website, where you can go to see your bill,
pay your bill, change your service plan, etc.  They also have a VRU
(Voice Reponse Unit) that offers the same content via a telephone
call.  *Lots* of custom code.  *Lots* of money.

Could you explain your application a bit more, so we understand the
scope of what you are working toward?

---

Clarification of Question by coolpal-ga on 31 Jul 2002 09:00 PDT

Yes, thats the idea. The voice to text and text to voice is relatively
new but is gaining popularity with time. We have a site that
recommends travel suggestions based on their preferences (like budget,
duration, location, type etc). Its a straight-forward database look-up
appplication right now, but we want to take it one step further. 
Please feel free to ask for any more clarifications.  Thanks.

---

Clarification of Question by coolpal-ga on 31 Jul 2002 09:03 PDT

I can already parse the natural language once it is in text form and
scan it for keywords for a database query, but I have not done any
investigation of
the tools and technologies to go back and forth between speech and
text.

---

Clarification of Question by coolpal-ga on 31 Jul 2002 09:08 PDT

what I need is to be able to speak into something and have text
generated, and submit text to something and have speech generated --
it doesn't need to be the actual advice engine.  Hope that helps.
Thanks.

---

Request for Question Clarification by blader-ga on 31 Jul 2002 15:48 PDT

Dear coolpal:

There are consumer software products that either translate written
speech to text, called voice recognition software (such as Dragon
Naturally Speaking), or synthesize voice from text (various), but none
that do both, as far as I know. Would a list of both of types of
software be what you are looking for?

Best Regards,
blader-ga

---

Clarification of Question by coolpal-ga on 31 Jul 2002 18:07 PDT

dear blader-ga

"Would a list of both of types of
software be what you are looking for?" - No, it would not.  
The question is pretty much clear. Comments by voyager-ga, lot-ga and
siliconsamurai-ga are all in the right direction, but they are not
answers to the question. I think they know what I'm looking for, but
don't have the answer as yet... If you go through the posted comments,
you might get an idea of what I'm looking for.
Thanks a lot, and please ask for any further clarifications required.

Hello,
you may find this source useful to find the voice portal application
Google web directory: Computers > Speech Technology > Voice Portals 
http://directory.google.com/Top/Computers/Speech_Technology/Voice_Portals/

Polaris Active article seems to have an easy to implement  ASP
solution
http://www.polarisinteractive.com/vp/products_faq.html

"Q:	How long does it take to voice-enable a Web site?
A:	In most cases, voice-enabling a Web site can be done in a matter of
hours or days."
contrasted against Intel's source which seems to be a much more
expensive and harder to implement solution
http://www.intel.com/network/csp/products/6928web.htm
PDF file: http://www.intel.com/eBusiness/pdf/busstrat/telco/wp013106.pdf
I've tried some implement from eckoh.com and the voice recognition
system is inoperable in noisy environments as the system gets
confused, street, train, car, TV all confuse the voice commands.
However there is keypad override which helps in such circumstances,
and would be wise to implement.
regards
lot-ga

I just got a Beta copy of Microsoft's .NET Speech which looks like it
might meet some or all of your needs. You can find information about
.NET and the new .NET Speech Beta tools at the Microsoft Web site. I'm
not attempting to provide an answer because this is a Beta product and
therefore development costs are difficult to estimate but I am posting
this information in the hope that it may provide a useful lead if no
existing products meet your needs.

Hi!

I think you should at least have a quick look at VoiceXML (
http://www.w3.org/Voice/ ) from the W3C. I didn't do an in-depth
study, but had a quick look into it a while ago and it seemed pretty
usable.

voyager-ga

You just caught me as I was about to hop offline for a minute, but I
thought that I would offer this small iota of advise:

Such products are available, although they would have to be altered
for web use, and I have used a product that seems to fit your criteria
in communicating by the phone with my friend, who is deaf.  However,
the hardware to do this, as far as I understand, is hightly
specialized and expensive.  Less expensive versions may be found in
the afore-mentioned Dragon technology.  My younger brother also has a
program that has a little purple monkey he can talk to or type to, and
it repeats what he types or talks to him.  I believe its name is
Bonzai Buddy, or something like that.  Although certainly not what you
are looking for, the mere presence of such a things such as what my
brother uses suggests that there is a product out there that fits your
criteria.  It would be extremely useful and I would be amazed if it
isn't already available.  I will try to research it more and will try
to pop back in and give you more details and specifics.  Good luck and
hope to be of more help really soon.

Dexterpexter