4 ms·

Ask HN: Community Validated User Profiles

1 points by alanthonyc 17y ago

I'm working on a community based web app and have been pondering ways to identify spam accounts. Aside from the usual suspects during the registration process (captchas, honeypots, etc.), I'm also thinking of adding a "feature" where someone in the user community validates or vouches for a new member. The username of the user who vouched for you would be stored as part of your registration information. There would functionally be no difference between a validated and a non-validated member aside from being unable to vote on some community related activities - in which case it would be a simple matter for someone in that community to vouch for him/her. Any thoughts on or criticisms of this idea would be appreciated. thanks

notaddicted 17y ago
http://en.wikipedia.org/wiki/Web_of_trust#Web_of_trust_problems http://en.wikipedia.org/wiki/Web_of_trust#Web_of_trust_probl...
shorbaji 17y ago
Well, this would rule out anonymous contribution to "community-related activities". Not sure how, but this may be an issue.
bdmac97 17y ago
Are you sure you're not going overboard with that? Is spam that much of a problem for your site that this is necessary? I've never been a big fan of throwing up too many roadblocks to real people to keep out the minority of spammers. If you already have captchas and honeypots and JS updated "type this [value]" fields in your signup process, do you really need more???
TallGuyShort 17y ago
I think you might be solving the wrong problem here. Someone could very easily set up an account and have it vouch for a bunch of spam accounts before you figured anything out. To solve that problem you'd have to limit "vouching" privileges to a few trusted accounts, and then you're not getting much of a benefit, IMO.
alanthonyc 17y ago
Hi everyone, Thanks for all your feedback. Your comments are well appreciated. I had come up with many of the same objections as well, but didn't have anyone to bounce the basic idea off. (I guess this is where having that co-founder comes in handy, as pg suggests.) @notaddicted - that was a good read, thanks. @TallGuyShort - your perfectly valid point would be solved by a solution that brings up other problems (i.e. who would be the trusted authority for validating accounts) as per notaddicted's link In the end, some subtler means of identifying, managing and controlling the undesirable accounts would be better. I'm thinking along the lines of what pg is doing with this site with karma thresholds, etc. Thanks again for all your insight.

1 points by alanthonyc 17y ago

notaddicted 17y ago

http://en.wikipedia.org/wiki/Web_of_trust#Web_of_trust_problems http://en.wikipedia.org/wiki/Web_of_trust#Web_of_trust_probl...

shorbaji 17y ago

Well, this would rule out anonymous contribution to "community-related activities". Not sure how, but this may be an issue.

bdmac97 17y ago

Are you sure you're not going overboard with that? Is spam that much of a problem for your site that this is necessary? I've never been a big fan of throwing up too many roadblocks to real people to keep out the minority of spammers. If you already have captchas and honeypots and JS updated "type this [value]" fields in your signup process, do you really need more???

TallGuyShort 17y ago

I think you might be solving the wrong problem here. Someone could very easily set up an account and have it vouch for a bunch of spam accounts before you figured anything out. To solve that problem you'd have to limit "vouching" privileges to a few trusted accounts, and then you're not getting much of a benefit, IMO.

alanthonyc 17y ago

Hi everyone, Thanks for all your feedback. Your comments are well appreciated. I had come up with many of the same objections as well, but didn't have anyone to bounce the basic idea off. (I guess this is where having that co-founder comes in handy, as pg suggests.) @notaddicted - that was a good read, thanks. @TallGuyShort - your perfectly valid point would be solved by a solution that brings up other problems (i.e. who would be the trusted authority for validating accounts) as per notaddicted's link In the end, some subtler means of identifying, managing and controlling the undesirable accounts would be better. I'm thinking along the lines of what pg is doing with this site with karma thresholds, etc. Thanks again for all your insight.