This is something I think a lot of parents would want for their children. However, I've found that parents who are on the Internet a lot know about chat rooms and are responsible with their children; the parents that need this the most are the ones who don't use a computer often and don't realize the danger that their children face online. So if you're going to market this to the masses, it should definitely be in an offline form; on TV, on the radio, or in print, you'll reach a much broader audience that might not be so Net-savvy.

On the same note, this product should be easy to use yet secure, so that the most computer-illiterate parents can use it, but the sneakiest children have trouble disabling it. A lot of products out there can be killed with a simple trip to the Task Manager; to really protect computer-friendly children, it would need to be elusive.

As per Slauncha's comments, this would almost have to be bundled in with a bunch of other security/privacy tools. Maybe just code it and sell it to AOL as part of their feature set.

The sneakiest children aren't going to disable it. If they want to put the info in there bad enough, they'll use 1337 or whatever they need. It's the inadvertent, occasional slip up that this prevents...

Another way of managing privacy is to establish a trust network for the children's peers. Give the parents visibility into their childrens peer group. Parents grant the peers trust levels based on their own criteria. Monitoring policies can be defined for each trust level.

I think the goal is important and interesting. The market of concerned parents is huge. I think it is important children learn social skills, but it's difficult to let a program take care of that. The trouble is that the children ofter are more tech savvy than their parents.

You might be on to something, but I don't see this working, so I'm not voting for it.

Julius, I don't think we have to be 100% effective. Assume 10% of kids are savvy enough to get around the controls, that still means 90% of the kids are getting protected.

Yes! Would need a very smart parser to ensure that very smart kids don't encode the private information to evade controls.
I agree with this being best deployed as a plugin to the IM program. One model wouold be to reach agreements with the various IM vendors, but what if the kid downloads something? Perhaps have a standalone plugin that "infects" any IM program at startup (pie in the sky)

This is a product that needs to be developed. In fact, as a parent I wouldn't even care if the engine blocked appropriate content sometimes too.

This could be integrated well into an open source app like Trillian. However, I'd say that less than .1% of kids use Trillian(or other open source IM programs). In order for this to be effective it has to integrate with AOL, MSN or Yahoo's IM programs. Are there any existing programs that show this type of integration is possible? I haven't heard or seen of any.

Hey poet,

Sounds like you're technical like me because you're already solving the difficult problems! That's cool, but I don't think it's necessary at this point.

Remember the Cambrian approach is to get the idea out there and either thrive or wither in the market place. Lets solve the simplest problem first. If the market accepts our inital offering and demands a more robust solution then we'll address the harder (and more expensive) problems.

So we could do a simple parser first and get the basic functionality worked out. Let the market decide what's needed after that.

More thoughts on the potential of this idea...

Give the parents visibility into who their kids chat with, how long are they chat, and what kind of language are they using.

The software might do the following:
- Maintain a list of friends the child chats with
- Maintain a set of trust categories that friends are assigned to: E.g. friends the parent has met and trusts; friends from school the parent hasn't met; complete strangers; etc.
- Maintain a list of alarm words. The list could be downloaded from a service that Cambrian provides.
- Capture information about friends: chat frequency, chat duration, alarm words used, etc
Trigger alerts when some threshold is crossed

In operation the software could utilize the features to provide these benefits:
- know who your kids are chatting with
- keep track of how much chat time is being used
- inform parents when their kid is chatting with a new friend
- inform parents of inappropriate chats
flag potential bullying, violence, or suicide

Obviously there are privacy and trust issues between the parent and child that should be treated with deep respect.

I've submitted a similiar "kid chat" idea based on kid safety, music, and design. Maybe we can brainstorm.
http://www.cambrianh...er/ideas-id/klDzK8e/

I normally don't like to post It's been done, but in this case:

http://www.netnanny....p/page?sb=artgetmost

Net Nanny is arguably the best recognised brand for computer protection, and they have this feature as part of their service.

Damn! This is getting hard! I like Spoil My Spouse too! Then Aidan and others note that both ideas have been done already! Seems to me that it dooms both ideas.

However, how important is being first to market? Is it possible that ideas developed under Cambrian House will be developed cheaper/faster?

How well entrenched are the existing apps/web sites?

I think the main problem with attempts to do this is that kids know how to disable programs on the parent's computer.

The best way I can think of to do this is as a service installed on the ISP's server itself - so when you connect to the internet and have your username/password authenticated, it checks your parent settings on the server.

Since all communication has to go through that server to the rest of the internet, and since kids don't have any access to the setup on that server, no matter what program they use to chat, - even if they are playing an online video game - the text would get filtered.

You could even have a way for parents to log in and see how many alerts/prevented information events occured, and probably decifer what program was used to chat with by having the component examine the TCP stream's format.

It is still possible to get around this - using a encryption is very easy (a program that you say, type a sentence into and it converts it to junk, and the person on the other side uses the same program to convert it into english).

Still, Internet Service Providers often advertise what special anti-virus and anti spam software runs on their servers as a benefit to their customers, and having some sort of "safety family filter" logo to show they offer this service would give them a competitive edge against providers that don't.

I think as long as it wasn't buggy most providers would happily pay to have this component installed for their customer's use.

I wanna comment on the "kids can just use leet" idea. Yes. So can we. If we add this into the features, this will set this product apart from the other services out there.

Have this filter trap for all leet cypher variations of the information that's being protected. ALSO: If it detects information being entered, DO SOMETHING ABOUT IT. Don't just try to block the kid, pop up something explaining to the kid what was blocked, and what the consequences could have been if it wasn't blocked. Let the parent add stuff to the end, sure, but we would have to include the kid in this, helping them make better choices. Otherwise it'll look like just another restriction in their power struggle with their parents.

Also with teh 1337z0rz, crowdsource this part, I'm serious. Have translations of cypher and terms so parents will know what their kids are talking about when they see what their kids have been up to. Let the community help keep this up-to-date; I know I've personally developed my own cypher and posted it based on researching others', so I would submit my alphabet key to the community. Algorithms would do their best, and parents could submit to the community items that didn't quite make it into our filter. Perhaps offer message boards where parents could share their experiences and advice.

This would also add mass market appeal. Many parents are scared of this whole other language they don't understand. We could help them out, and show them that usually, it's nothing to be ascared of. With this, we may also want to explain to the kids why we think it's important they not hide from their parents what they're up to... maybe a message board for them, too. If we could find a way to get the kids to not hate us, we'd be golden.

It's impossible to check for all possible maskings. What if they insert a period, like "3 Fu.bar St", or even mess up the form entirely like "the street called Fubar, house number 3"?

The idea is nice, but I just don't see it as being practical. The solution to this is to work *with* the kids (via education), not against them.

Somethings I don't get. Is it supposed to be hidden from the child? If its supposed to be hidden, how can you keep it that way when the child's chat partner will see when some information has been censored? If it isn't hidden from the child, then there are an infinite number of alternate spellings, codes, and phonetic spellings that the child will find a way to circumvent the filter.

Just an idea:

What if it blocked nothing at all?

What if all it did was report suspicious stuff to the parents?

That way - the kid could not learn what is/not is passable and work out ways to beat it. They would never know if they are getting away with it until they were caught.

I don't think it really helps parenting to give them what they could misuse as a "fire and forget" safety feature when the times the system goes off are the times the parent really needs to talk to the kid.

Besides which, they can't even fill out an online school registration form without giving out a home address - you can filter content but not context...if it just reported the content with enough other text that the parent could review the context themselves, they could decide what instances are good and bad and when to talk to their kids.

I think the only way to be completely safe is to install a monitoring program that saves the whole conversation and then allow the parent to view it - either in a database or via an email. The parent reads the content and decides whether the information is acceptable or not.

I agree that it's not possible to create an algorithm that can detect all possible ways in which an address might be mangled. Look how hard it is for spam filtering software to monitor what's in the Subject field in an email.

While it may be possible to configure a computer so that it does not allow disabling of the software, this is by no means easy. And kids could still use computers in other people's homes, internet cafes, etc.

Short of watching over everything your child does, the key is education and trust.