-----Original Message----- From: Alex Broens [mailto:surbl@alexb.ch] Sent: Friday, October 08, 2004 12:51 PM To: SURBL Discussion list Subject: Re: [SURBL-Discuss] Revised DMOZ data, got Wikipedia domains too
Bill Landry wrote:
----- Original Message ----- From: "Alex Broens" surbl@alexb.ch
Jeff Chan wrote:
http://spamcheck.freeapp.net/whitelists/wikipedia-dmoz.srt
Please also take a look at these blocklist hits (potential FPs) and share what you think:
http://spamcheck.freeapp.net/whitelists/wikipedia-dmoz-blocklis t.summed.txt
Would there be many FNs (missed spams) if we whitelisted all of these? In other words are these all truly False Positives? If not, which ones do you feel are true spammers and why.
probably not a new idea, but why not run a "wl.surbl.org"
with all the
whitelisted domains and ppl can choose to use it or not.
I like this idea! Whitelist the most commonly used 1,000 or
so domains, and
then create a wl.surbl.org for the rest of the
wikipedia-dmoz domains.
WOW... Bill didn't bark at me this time.
my point is the following:
take for example "angelfire. com". This domain may have legitimate users but my user base would NEVER have contact with anybody hosting a site or anything there. If they support spam, list them, put pressure on them to stop supporting spammers, bla, bla, bla. I wouldn't appreciate it being whitelisted as then if there's abuse, and it does get blacklisted, there's no pressure on the domain holder to clean up.
I'm hoping UC list will help with that. Anyone who wishes to use it can add a minor score to spam for UC hits. This would put at least some sort of preasure for these grey hats to do something.
As I imagine we're fighting spam here, not just filtering, I have a certain difficulty understanding why the world is crying for whitelisting instead of putting pressure on so called whitehats who support abuse for a lifetime.
I completely agree, but when working with SURBL I can't use this mindset. I have to go by Jeff's vision. However with UC its a whole different ball game.
as Chris said, you could make whitelisting a lifetime task. I believe the better approach would be to decrease potential FP's by increasing the reporting QUALITY !!!!!!!!
Alex
//Are we fighting Spam or working for Messagelabs & Co. for free? //
OUCH!!! I hear ya.....but OUCH! ;)
--Chris
Chris Santerre wrote to 'surbl@alexb.ch' and 'SURBL Discussion list':
I wouldn't appreciate it being whitelisted as then if there's abuse, and it does get blacklisted, there's no pressure on the domain holder to clean up.
I'm hoping UC list will help with that. Anyone who wishes to use it can add a minor score to spam for UC hits. This would put at least some sort of preasure for these grey hats to do something.
Ezzactly. Even if it's worth 0.1 points in SA (or something), the stigma of having one's domain listed in some sort of bad "list" might be enough to promote action, provided the accuracy of UC is quite good. (I.e., we definitely still don't want to list everything :-) "Quite good" in that case probably means < 1% FP, but I think we can do even better than that with a few clever tricks.
Personally, I'm hoping UC will be worth more than 0.1 even with a somewhat higher FP rate, since there will be little or no overlap with other SURBL tests, but that'll depend on some mass-checks. :-)
- Ryan
----- Original Message ----- From: "Ryan Thompson" ryan@sasknow.com
I wouldn't appreciate it being whitelisted as then if there's abuse, and it does get blacklisted, there's no pressure on the domain holder to clean up.
I'm hoping UC list will help with that. Anyone who wishes to use it can add a minor score to spam for UC hits. This would put at least some sort of preasure for these grey hats to do something.
Ezzactly. Even if it's worth 0.1 points in SA (or something), the stigma of having one's domain listed in some sort of bad "list" might be enough to promote action, provided the accuracy of UC is quite good. (I.e., we definitely still don't want to list everything :-) "Quite good" in that case probably means < 1% FP, but I think we can do even better than that with a few clever tricks.
Personally, I'm hoping UC will be worth more than 0.1 even with a somewhat higher FP rate, since there will be little or no overlap with other SURBL tests, but that'll depend on some mass-checks. :-)
I have been tracking my hits from UC, and although the hit rate if fairly low compared to most of the SURBL lists, it has been giving me good results and hitting things I would consider spam. And as such, I have the test currently weighted at 0.5.
Bill
-
Bill Landry -
Chris Santerre -
Ryan Thompson