#14 New member authentication through voice recording

[mark]

BACKGROUND

The objective of this feature has to do with the new-member registration identity checking. There are 2 conflicting demands that are facing all the users of similar social networks.

On the one hand, everybody wants to the registration process to be very simple and painless. New members want to provide only some absolute minimum amount of private information when they sign up for any new portal, network or club membership. We all want to minimize the hassle. I sure know that I do.

On the other hand, if the registration identity checks are very loose, then there is complete impunity and anarchy in pretending to be somebody else, stealing somebody's identity and possibly even defacing somebody's reputation behind their back. Or pretending to be a famous actor, when in reality a marketing department of a movie studio is promoting a movie. Or promoting a new car model. Or an actor is hired to create some devious promotional marketing scheme - like the old lonelygirl15 story. (MySpace is especially plagued by the problems of deceptive new member registrations).

Here's our attempt at a relatively balanced solution: In our registration process, we have introduced a small second-order hurdle we call "authentication". Authentication is something of a conscientious promise and declaration that I am really declaring myself to be a true person. One approach that might be quite powerful, yet simple, could be a "voice signature". I do not believe it has been done before, and it could be quite an interesting feature. Also, it should be fairly easy for a user to do. Almost everybody has a microphone nowadays, with ubiquity of Skype.

Here is how it could work, for example: In the final step of the registration, we would ask the new member to read into his/her microphone the previously submitted user data, in a short statement, as follows: "My name is John Doe, I am male, 26-year old and I live in Springfield, California. I declare this statement to be true." That is just an example. We would have to work out the details a little later.

The audio (hmm.. maybe even optionally audio+video, if available at the client) from the new member's microphone would be streamed to our server and saved there for later playback by a member of our GreatCOW team. Frankly, there is not much we could do with it after listening. The only thing we could detect is if somebody is trying to play a stupid prank. Then we could refuse to authenticate the person. Otherwise, the member would be authenticated.

Of course there is no guarantee at all that this feature will eliminate fake registrations. That is not the point. However, it would make possible pranksters or fraudsters think twice about doing things which are wrong. Again, we should phrase the statement so that the new members know that they are leaving a permanent "voice signature" on our server. This may or may not be enforceable in the court of law. I have no idea, I am sure it has never been tested in the court of law

There is another exciting technical challenge to this feature. Originally, we wanted to use Flash Communication server from Macromedia (now Adobe). We do owe licenses to this commercial (closed) software. However, using it would not be in the spirit of open source. And I do not think we can bundle closed source with open source in the same GPL application. At least I certainly would not like it. The question is: can we implement this kind of feature without using proprietary solution like Flash Communication server. If there are any ideas, then we would be excited to here what you gentlemen and ladies have to suggest.

But Flash player client is great for any multimedia streaming applications. It is good, and it is very user friendly, and almost every user has it in their browser. ... So I do not think we can escape into some Java Applet stuff. There is unfortunately much resistance to that and the penetration in the browser is not so good. The downloads and installation are not simple and can be enormous.

This new feature item specification and requirement details will need to be further discussed and reviewed. But it is some hot stuff. (And we will definitely reuse this voice recording capability in any fun places on the GreatCOW ... I am thinking of an instant voice-messaging feature ... instead of conventional instant messenger. That would be really cool).

Cheers,
Mark

[mark]

Logged In: YES
user_id=1691791
Originator: YES

We have heard about the Open Laszlo project before. And now Jay has mentioned it too. Perhaps we need to investigate it and see if it could meet the requirements as specified in this task item: helping record, stream and save audio/video to our server? Would it work as a replacement -- so that we can avoid tying ourselves to proprietary FlashCom server?

Also I read a little bit about another promising OS technology called Red 5: http://osflash.org/red5

Adam, can you evaluate?

And Below I'm copying and pasting email from Jay:

-----Original Message-----
From: jay
Sent: Thursday, February 22, 2007 6:37 AM
To: mark-jj; Adam
Subject: open laszlo

Hi guys,

A friend of mine has been doing amazing things with open laszlo, and it
might be useful to us for some of the flash based tasks:
http://www.openlaszlo.org/

Take a look.
j.

[rbrocato]

Logged In: YES
user_id=1731830
Originator: NO

Hi Mark,

I would like to work on this, if I may. But, I don't want to start working on it if you're going to give it to someone else while I'm working on it because that person got done first. I will work on anything else you need done, if you wish.

I'm somewhat new and old to Java. New in the sense that I have been re-studying Java recently. Old in the sense that I haven't programmed in Java in about 8 years. A lot has changed since then. Anyway, How can I get a project that will be a learning expirience for me that no one else is working on at the same time?

Robert Brocato

[mark]

Logged In: YES
user_id=1691791
Originator: YES

Hi Robert. We definitely appreciate your interest. But let's wait a little with this task assignment. I think what we need is to completely review our entire web application and to redesign it to make it more modular if necessary. This feature would likely become its very own module. However, I think we need to completely understand what would be the architecture redesign implications, and then establish some design guidelines for all the team developers and members working on this fine project... We are working on it now with Adam, Jay and others. We are conversing.

Feel free to participate and add your thoughts to the most recent discussions going on in the forum, and the beginning of the new task that I wrote and started working on today: [ 1671606 ] Review high-level system architecture and modularity.

To answer your question. I think there is no 100% guarantee in the open source community that it will never ever happen. After all, the source code is completely open, so anybody can take it, start playing with it and run on their own developing their own things. However, in practice, usually it does not happen. Once you declare your interest here, we will understand that you are going to be the primary developer and the community honors the basic rules of work division. However, communication back with the community would be very important. We would need to know that you are truly committed -- primarily we would need to see some periodic work progress updates to see that things are moving forward at a reasonable pace.

Cheers,
Mark

P.S. You should also know that Adam has been evaluating some OS tools that we would use for this feature. Primarily Red5, in his task: [ 1669138 ] Evaluate substitutes to Flash Technologies