MW4D bi-monthly teleconference

minutes approval

Last meeting minutes: http://www.w3.org/2008/11/03-mw4d-minutes.html

<jeffs> +1

<steph> agreed

Action

Feedback on Nokia work on curriculum development in Africa from Lauri not yet

available. Action open

Steph keeps the action open regarding profile of participants

Steph: on scheduling mtgs

Presentation Voikiosk by Arun Kumar

<steph> http://www.w3.org/2008/11/Voikiosk-17Nov2008.PDF

Arun: on the slide presentation

15 to 20 minutes presentation

<jeffs> like the idea of "Information Literacy" rather than "Computer Literacy"

<steph> me too, very interesting concept

<jeffs> what level of local tech is needed to support a voicesite? voicegen software?

Would regional accents be a problem for voice recognition systems?

+1

<steph> steph question: how to update a homepage (author recognition)

<jeffs> jeffs question: what level of local tech is needed to support a voicesite? same question for voicegen software?

My phone in is weak for voice transmission right now

<NChevrollier> any better?

<steph> sounds good

A voice site would be very useful for our current project in Liberia

<steph> question: what is required to run the service, is the software free/open source ?

you did tell me in Sao Paulo.. :)

<jeffs> question 2 from jeffs: what is accuracy/error rate for this in field tests?

<Renjish> question: is there a translation facility to access the same site as a normal web page?

<NChevrollier> question: from first pilots, what types of services are the most promissing? who is the most likely to use this system?

The system would be usefull for the http://www.questionbox.org/

Arun: opening to questions

Steph: open the floor for questions which were collected during presentation

<jeffs> jeffs question: what level of local tech is needed to support a voicesite? same question for voicegen software?

Jeff: What technology is required?

<jeffs> the central server site... what level of local tech needed to support that?

Raphael: if anyone can help me catching what Arun says, thank you

<jeffs> what is cost of voice server s/w & h/w as vs normal website?

<steph> good question, depends on the technology

<jeffs> question 2 from jeffs: what is accuracy/error rate for this in field tests?

<Arun> In terms of h/w

<Arun> we use genesys voice browser

<Arun> tomact app server

Thank you

<Arun> IBM WVS Speech server

<jeffs> so a fair bit more tech needed locally & tech support skills

http://www.questionbox.org/

<steph> zakim ack raphael

Arun: inputs and outputs are taken by voice

<steph> renjish: translation, multi channel access

<steph> to normal web access

<steph> Arun: we are trying to work on it

Arun: support to people wanting to create their own applications

tools for these people

Arun: generic voicesite can be given and then later tailored to specific needs

<steph> arun: there are templates

Steph: voice browser includes speech recognition engine?

Arun: yes

Steph: any test with simpler voice browsers without voice recogn.?

<Renjish> This is why I feel that speech processing may not be matured enough to handle all kinds of nuances

Raphael: missing what Aruns answers...

sorry

<Renjish> :)

Steph: is the system completely integrated? is it modular?

Arun: not in release stage yet

working on it

Nicolas: who is using the system, in the pilot(s)?

Arun: being run by farmers

used, sorrx

run by a foundation?

<jeffs> question 2 from jeffs: what is accuracy/error rate for this in field tests?

Arun: no real study on it yet

most of the usage is yes/no answers

Jeff: You said it is preferable to SMS. SMS is preferred for the moment, do you have more arguments for voice?

<Renjish> literacy can be an issue in SMS in my opinion

<jeffs> so discomfort is about literacy issue, not technology?

<steph> ubiquitously available, literacy

<steph> ability to get a free number

Steph: voice is as ubiquitous as SMS, or more

on existing phones

especially on fixed phones

<jeffs> thinking about PRC experience with emergence of SMS-only accounts, not happening in India?

<Renjish> we can do reverse billing for pull SMS

Missed Arun points again. Anyone can sum up?

thanks

yes

thanks Arun

<Arun> We intend to expand since the foundation is very excited about this.

<Arun> They want to expand not only in terms of villages but also services.

<steph> renjish:scalability issues ?

Renjish: Will it be plug and play or more complex

<steph> Arun: dialect issues will be managed

<jeffs> what action do we wish to take on this report?

<steph> i see that in the voice apps section it should appears in the lists of tools/solution

I will check interest for such a solution in Monrovia in three days.

<Lauri> We should add this Voice-WEB-technology as one option to our list of "working solutions" to offer services.

<Arun> We have restricted the amount of voice input that needs to be recognized

Rapha�l: I can engage to give feedback to the group, during our next meeting.

<steph> ACTION: Arun to make a summary of voikiosk [recorded in http://www.w3.org/2008/11/17-mw4d-minutes.html#action01]

Steph: next call: Dec. 1

<steph> ACTION: stephane to start mail discussion on the organization of the work [recorded in http://www.w3.org/2008/11/17-mw4d-minutes.html#action02]

<steph> http://www.worldbank.org/edevelopment/live

<steph> Enabling Mobile Transformation in Africa – the discussion will focus on the Next-Generation of Mobile Services for Government, Health and the Financial Sector.

Seph: session organized by the World Bank tomorrow

Thank you

<Renjish> Thanks

<Renjish> bye