Send Me Your Spam

[MstrBob]

Hey everyone, I've got a bit of an odd request. I am examining filtering and prevention methods for comment spam. I'm hoping to find the most efficient methods, and see where it can be improved and such. But in order to be effective, I need a large and diverse sample of comment spam. Now we've all seen it before, and it's out there. I just need to collect it. So I ask you guys, if say you run website where you've got comment spam, can you send it my way? You can email me via the forum software (Click for my profile and the forum has a form which you can send me email to) or email me directly if you have my email address. Doesn't matter what format it's in, as long as it's plain text. Thank you kindly.

[the tree]

Just comment spam? No e-mail spam?

[rhsunderground]

i love this stuff

[JPnyc]

For everyone information, we're moving to have Bob committed.

[NogDog]

filter out any comment that has no capital letters and which has a ratio of letters to punctuation greater than 100 to 1 or which has more than one occurence of the word teh

;)

[rhsunderground]

^^^Spam

[David Harrison]

On my site I hold any comment for moderation if it has more than 5 links in it. And also, any comments from someone who does not have any approved comments already get held for moderation. Once they have one comment approved, all their other comments appear on the site immediately (unless they have more than 5 links in them).

So far only 1 spam comment has ever made it onto my site, because immediately after that I put up some spam protection. :p

Since Christmas I've filtered out 110 spam comments. They seem to be dying down a bit now, maybe they've realised that NONE of them are getting through.

However it does mean that you have to check them all manually, though it's usually pretty easy to see which ones are spam and which aren't, unless you get thousands of spam comments.

[MstrBob]

Hey guys, sorry I would have responded sooner but my internet service has been out for the past two days. :eek:

Yes, I'm looking for comment spam. Because E-mail spam and Comment Spam are different enough that your common email spam techniques won't be as effective. From how they are worded, to detecting the differences between ham and spam, I need comment spam to be on the ball with testing.

@Dave,

You see, exactly what I want to test is what is the most effective and efficient method. Does simple link moderation due the trick, or is bayesian filtering neccessary? How many false positives result from each one, ect. I want to get some real numbers down on this mess.

[Cinderella]

Is this an attempt to have us commit you?

[MstrBob]

Gah, internet cut out again today. Time Warner is really irritating me now...

Anyway, thank you to those of you who have helped me out thus far. Keep it coming though, the more the merrier I guess...

[Jona]

I already sent you mine, Bob. Do you still want more? I deleted 26 yesterday.

If there are two or more links in a comment on my blog (http://www.slightlyremarkable.com/) (shameless plug), they automatically are placed in the moderation area for me to decide whether they are spam or not. However, prior to this evaluation, Akismet (http://www.akismet.com/) decides whether it is comment spam or not; 99% of the time, Akismet picks up comment spam. Out of over 340 spam comments that I’ve had since I installed it, Akismet has missed 3 spam comments. It’s the most effective comment spam combative software I’ve used (and I tried lots of different software). I don’t know if it uses whitelists or blacklists or anything, though.

[MstrBob]

Sure, if you would. I'm still accepting spam comments. The more I have, the better I can determine the effectiveness of different filtering methods.

Though I do agree with you. I use Akismet myself and haven't had any spam slip through. But spam is annoying enough to warrant, at least for me, a bit of research into the effectiveness of filters.