3 ms·
Very true. The best you can do is allow the post to go thru and make it appear to the spammer that they have succeeded. Let them view their post on the homepa
by jasonkester 18y ago
Very true.
The best you can do is allow the post to go thru and make it appear to the spammer that they have succeeded. Let them view their post on the homepage and go away with a nice warm feeling of satisfaction, secure in the knowledge that they don't need to improve their algorithm at all for this site.
It's pretty straightforward to modify your display mechanism to include spam posts if and only if they originate from the same IP address as the viewing client.
I implemented this on Blogabond, and it had two positive effects: My Spam corpus is growing at a faster rate (thus making it more effective), and the sophistication level of the average attack has dropped sharply.
- philh 18y agoThe problem is detecting the spam posts in the first place. Your idea would be incompatible with the proposed method, because you don't want to do the same thing to legitimate new users.
- jasonkester 18y agoCertainly, HN must have a bayesian classifier looking at the content of its submissions and deciding whether they are "Hacker Newsworthy". By now, the site must have sizeable corpi (corpuses?) of good articles and spam links to use in such an assessment. Since this site was built by Mr Bayesian himself, it never occurred to me that it might have weak spam filtering.
- dagobart 18y ago> The best you can do is allow the post to go thru and > make it appear to the spammer that they have succeeded. I read the cons on this suggest, but what about turning the approach just around: Keep the spam in by default, just the folks with higher karma get the spam filtered out? That way, the successfully submitted spam could not only be verified from the original source but also from every other node of, say, a bot net.