Slashdot Mirror
Two Spam Filters 10 Times As Accurate As Humans
Nuclear Elephant writes "The authors of two spam filters, CRM114 and DSPAM, announced recently
that their filters have achieved accuracy rates ten times better than a human is capable of. Based on a study by Bill Yerazunis of CRM114, the average human is only 99.84% accurate. Both filters are reporting to have reached accuracy levels between 99.983% and 99.984% (1 misclassification in 6250 messages) using completely different approaches (CRM114 touts Markovan, while DSPAM implements a Dolby-type noise reduction algorithm called Dobly). If you're looking for a way to rid spam from your inbox, roll on over to one of these authors' websites."
I'm sorry, Dave... That Nigerian guy looks suspicious and I can't let you send him money.
------- "A true friend stabs you in the front." -Eliot
Comment removed based on user account deletion
Once Email Spam is eliminated, then IM spam will begin...
Jeff
I presume they mean more accurate than a human that was only looking at the subject line? I fail to see how someone could misclassify an email after they'd already opened it unless it was some kind of marathon testing, which would be totally unrepresentative of any real life situation. Once you're getting 6,000 messages, it's time to reach for "Delete All" and change your address, methinks
can this be used with Spamassasin, or is a stand alone program? Does it need something like Amasis to run?
CB
free ipod and free gmail!
If your email is indistuinguishable from spam by a human, perhaps the problem isn't the receiver. It's the sender.
Forgive me if I don't feel any pity that some moron's email gets filtered to the junk bin because I couldn't discern it from spam.
I have been pwned because my
Just enter a valid email address, and hit submit!
I haven't been 100% accurate.
I received an email from my sister-in-law from her work, and the address looked suspicious (one of those weird-looking "letter and number" jumbles.
I deleted it. It happens.
Well, it certainly sounds better than the pay-per-email "postage" idea. If postage hasn't stopped snail spam, why would it stop e-mail spam?
Human=99.84
.16% and the machine only missues .016% hence the machine is 10 times better.
New proggie=99.984
So the human misses
I suppose it depends how you're defining spam. Perhaps the ultimate spam messages that don't get past them are capable of passing a turing test... hence fooling those gullible human recipients into thinking that it isn't even spam!
Fortunately, soon we will all be able to use the superhuman spam-detection capabilities of these filters to save us from ourselves. Imagine all of those pesky e-mails from your 'friends' getting caught by your spam filter before they even impinge upon your consciousness.
It'd be a wonderful world.
"We have to go forth and crush every world view that doesn't believe in tolerance and free speech." - David Brin
If you read the post, it quotes a study and says humans are only accurate 99.84% of the time.
:)
Kinda makes you wonder how they can know the filters are right though.
(please don't reply telling me how)
Probably used those same people who open viruses as test subjects.
I reached the conclusion of "two filters better than humans" by using two sequential filters:
server side spamassassin, and a couple of simple procmail recipes. They have kept almost all the SPAM away.
However, it is good to see such good techniques becoming available and we can hope to see them as straight forward usable tools.
So, when will mozilla/TB (or your favourite server side or client side filter) get them?
S
it's not that humans are not as accurate, it's that 1 in X times we really do want a mini camera or free porn. It is what seperates us from those cold, heartless machines.... mini cameras and porn....
Obviously you've never seen someone new to the internet sit in front of their computer. Lots of people don't know what popups are. Lots of people read some spam not knowing what it is. To these people, a computer is merely an interesting string of sensations.
Know what I like about atheists? I've yet to meet one that believes God is on their side.
I received an email from my sister-in-law from her work
Yeah, so did I. The subject line was "I want you so bad."
I deleted it. Turned out the message was genuine. I'll never forgive myself...
The coolest voice ever.
I'm also sure that Yahoo's "SpamGuard" was great when they first introduced it. Now, It catches roughly half of all the spam I get. Why? Because people have figured out how it works and taken advantage of it. The same will happen with any content-recognition-based spam software. In the extreme case, even if a piece of software were 100% accurate at saying "This piece of e-mail looks like spam," then spammers would just make their e-mails look exactly like e-mail from one of your buddies. How could software ever tell the difference between:
Hey, dude, check out this website I found. There are some hot naked chicks and stuff. Sweet.
Signed,
Your Buddy
and
Hey, dude, check out this website I found. There are some hot naked chicks and stuff. Sweet.
Signed,
SpamKiddy
Even a human can't tell the difference. The only real difference is who they're from.
Perhaps they mean that Human A is reading email intended for Human B and attempting to classify the email as spam or not spam. I wouldnt be surprised if a computer could do a better job at that sort of task. Besides Im sure Human B wouldnt want Human A reading that cyber sex chat log.
[*Bing* -- mail from VP of sales pops into my inbox. Subject: "Making money fast!"]
[*Bam* -- I hit delete, thinking "Stupid Spam!"]
Ahh, shit! Lookie, a human screwed up.
The filter would have actually examined the message and probably decided that it was legitimate.
It's really easy to design an effective solution when the problem is purely mechanical or natural. As long as you're working with spammers who don't adapt, you can slice through their shitstorms very effectively.
But when a single solution becomes mainstream, spammers will adapt to it. Bayesian filters tend to work very well, but now spammers are adding sprawls of randomly generated green-light text to offset the filter's score.
Google found an excellent way to rank websites, but then it became widespread enough that webmasters began to game the system it had created. It's been playing catch-up ever since.
Once the adversary begins to adapt, we lapse into the same cat-and-mouse game of technological barriers and counter-barriers that we've seen so many times before.
Quite simple:
With 10 messages (after automatic spam detection) humans are 100% accurate.
With 1,000 messages, (before automatic spam detection)
humans are less than 100% accurate.
The experiment was done on 5849 messages.
Remember; one thing computers are good at is doing boring things repeatedly.
No, humans are not 100%.
If you see a strange name in your inbox with an odd title, that might be a Nigerian businessman, or it might be your long lost Nigerian brother.
I recently tried to order a t-shirt from this guy for a band he used to be in. I found his band because we have the same (semi-uncommon) name. So, he got an email From: himself. I had to send him two emails because he deleted the first one assuming it was spam.
I ordered some RAM for my dad a while back. He gets 200 spam emails a day (email addy in resume & web page), and he deleted the confirmation email from the RAM vendor. The RAM never shipped, and it took us a week to figure out that there was a problem.
People make mistakes all the time. Why is this an unexpected result? People are jackasses. This should be obvious.
There are no trails. There are no trees out here.
I order all kinds of stuff online, wouldn't the receipt emails look like spam? My current spam solution is very simple:
1. display my email online as little as possible
2. use a number of addresses that all filter into one account, then filter by the sent-to address... this has turned up some VERY interesting results, for instance. I used dellorders@mydomain.com for an order from Dell, and NEVER used it or even typed it anywhere again, and started get spam about 6 months later, and I mean the nasty stuff, no just innocent stuff from Dell resellers...
3. i built a rudementary filter that looks for viagra,free,debt,enlarge, etc... if the sender is not in my address book, and the email contains these words, it is sent to a "check these out" folder...
How might a spam filter help me out without zapping confirmation type emails?
Cloud City Digital: DVD Production at its cheapest/finest
And if the study posted about is accruate, of those 1% that are left, you will (if you're a perfectly average person) accidentally delete 0.16% of good messages. Surely you've deleted a valid message by accident before? I do it regularily, deleting 25 spam messages with a single good one embedded in it when I just woke up before I had my coffee is not a good thing ;)
At the very least, if you were given the same data as these tests, that would be true. Consider if you *didn't* use popfile - how many spams would you be deleting every day, and how many good messages would be accidentally deleted? I know that if I had to manually delete the two or three hundred spams interspersed with good messages, my false-positive rate (the percentage of good mail I accidentally deleted) would skyrocket.
So just be glad you've got popfile. Not only do you not have to go through as much spam, but you're also more accurate while going through the little you must.
Barclay family motto:
Aut agere aut mori.
(Either action or death.)
Results of new spam filters cannot help but to be bogus... The true test of a filter is how well it works *after* all the spammers know how it works and try to circumvent it.
Enjoy science fiction? "Turing Evolved" - AI, Mecha, Androids and rail-gun battles. What more could you want?
A CRM114 plugin for SA is available, thanks to Devin Nate:
d =2 301
http://bugzilla.spamassassin.org/show_bug.cgi?i
client/server-side filtering does NOT solve the problem!
The biggest problem with spam is the invasion of third party computers on the Internet. The ILLEGAL activity spammers perpetrate by breaking into machines, forging headers and hijacking servers.
Any filtering method does not address this most serious problem, and even if you do not see any spam in your inbox, you're still paying for the bandwidth and system resources these spammers steal.
Stop with the filtering algorhythms and take some of that energy and contact your local Attorney General, DA and FBI and demand that they prosecute these people who are BREAKING THE LAW.
Don't worry, I can forward you the one she sent me. Sounds like the same email.
I remember when legal used to mean lawful, now it means some kind of loophole. - Leo Kessler
They are learning algorithms. For measuring their accuracy you have to assume that the data is correctly classified so you can see how they do.
The point is that humans also aren't perfect. Have a person classify 10000 emails and they will make a few mistakes. Point out those mistakes, and they will say "yes, I got that wrong it is an email from my wife reminding me to pick up milk and not a spam trying to sell me printer ink, I must have been day dreaming."
Just like if you give a person a document and say "find all the spelling errors" they will probably miss some. This is not because they have a different definition of how those words are spelt, it is because they made some mistakes.
For the training/testing data, some double checking needs to be done to find the mistakes the human classifying it almost certainly made.
It's a pretty normal situation in any machine learning application, you don't have to be perfect to be as good as a human - after all humans are only human.
Lots of people don't know what popups are.
Uh, sure they do. Popups - that's like those porn storms, isn't it? Some people say it only happens with IE and Windows, but I talked to my service provider and they told me 'just pull the power plug out of the wall when that happens'.
Easily fixed.
That actually makes humans much more accurate. We can eliminate many of the messages just by looking at the subject.
The further question is, if humans aren't as accurate as the computer, how are they measuring the accuracy at all? That is, how do they know that the 1 in 6250 messages is wrong, if a human, known to be inaccurate, was testing for accuracy?
Karma: It's all a bunch of tree-huggin' hippy crap!
Saying an algorithm is x% accurate is not sufficient, because there are two kinds of errors: false acceptance of spam, and false rejection of non-spam. Personally, I'd settle for 90% false acceptance if I knew the false reject rate was 100% rather than have a program that was 99% at both.
Also, I wonder how many people have actually looked at CRM114 and tried to use it.
The really interesting thing about CRM114 is the windowed polynomial hashing technique used although there's some evidence that it can work just as well (if not better) on a much smaller window of only two tokens. I'm hoping someone will do a full exploration of the idea for SpamAssassin's Bayes module.
Presumably they must use a superhuman who has 100.00% accuracy.
Karma: It's all a bunch of tree-huggin' hippy crap!
No, humans aren't 100% and yes, you can test for that. Try a thought experiment: fill a bin with 50,000 red balls and 50,000 blue balls. Ask a human to sort them all. The result probably won't be 100%, but you can still check the result and figure out how accurate the human is without relying on a superhuman ability to tell the balls apart. Same thing for spam: if you start with a known training set, you can test humans to see how well the spam is identified by manual sorting.
I'm not surprised a filter beat the human, considering the study used a sample of 5849 messages. As the sample size increases, the filter's accuray will increase, and the human's will decrease. Furthermore the higher the spam/real ration, the better the filter will do in comparison to a human trying to sort at a reasonable speed. The reason being humans tend to skim, and rairly actually read entire subjects, much less messages. Give a human 5000 messages and an hour and he will probably make some mistakes. On the other hand, in 10 messages, the human will probably be 100% correct. Most email filters rely on this already, as they tend to err on the side of caution. With the bulk of the spam taken out, it is not a burden to have the human check the iffy bits. Furthermore the type of email can mislead humans. A business-type email sent to someone's personal email is much more likely to be mistaken as spam, and vice versa. The main disadvantage of automated filtering is people generally have an idea of when a really important e-mail is going to come (the type that false positives are completely unacceptable) and who it will be from.
Congratulations, Mon Ami.
You have just unlocked the secret of virtually every news report that says "ten times more likely."
To get cancer. To have a heart attack. To suffer from the heartbreak of psoriasis. Whatever.
Yes, these numbers indicate "10 times better," and if you were to ask the reporter how likely am I to avoid cancer in both situations, these are the sorts of numbers he would show you.
Eat health food and your chance of having a heart attack is 99.984%. Eat too many donuts and your chance of having a heart attack is 99.983%, 10 times worse!
Always, always, always ask to see the raw numbers so that you know what "10 times worse" means.
Then ask if the numbers were collected by phone survey. If they were, throw them all away and have donut and a cup of coffee.
KFG
The post quotes "a study" which gives the 99.84% figure. In fact, the 99.84% figure is mentioned in the one paper as "the human author's measured accuracy as an antispam filter...on the first pass". This is what we who understand statistics call "nonsense". An individual human had an estimated accuracy of 99.84% when looking at one particular sample set of data, once. This is not a meaningful number, and it sure as heck ain't "a study".
I talked to my service provider and they told me 'just pull the power plug out of the wall when that happens'.
Ok, now the screen dimmed a little and I heard the hard drive spin down, but the pop ups are still a comin! Oh, and something about "battery level at 98%" or something.
I'm not a journalist, but I play one on slashdot
There goes my bussines idea. I wanted to start a bussines that puts humans in an eastern europe contry to sort corporate e-mail.
Now I have to think again about putting humans to decorticate sunflower seeds, it's cheper than all those machines.
The further question is, if humans aren't as accurate as the computer, how are they measuring the accuracy at all? That is, how do they know that the 1 in 6250 messages is wrong, if a human, known to be inaccurate, was testing for accuracy?
I believe that humans can be 100% accurate (or thereabouts) if they read the *ENTIRE* message, however that's exactly the point - if you have to read an entire message to tell that it's spam, the spam has succeeded.
Their number probably concerns how people can tell without reading the entire message whether or not the message is spam. My brother accidentally deleted a few messages I had sent to him, however if he had read them fully he would have known they were legit.
Cheers,
Justin
But the computer reads the entire message, so it's not really a fair comparison, is it? How many more lines of information was the computer allowed to look at to create its superior result?
Karma: It's all a bunch of tree-huggin' hippy crap!
You mean you've never noticed this before? Idiots are some of the happiest people I know.
ObKubrick: In 2001: A Space Odyssey, one of the pods was marked with the designation CRM-114. And in Clockwork Orange, Alex is injected with serum 114. I suppose CRM-114 is to Kubrick as THX1138 is to Lucas.
Dobly, on the other hand, is from This is Spinal Tap , a mispronounciation of "Dolby" by David St. Hubbins's girlfriend:
Not to mention that it probably avoids trademark infringement (though I wouldn't put it past Dolby Labs or Thomas Dolby to raise a stink).
Maj. Kong
Shoot, a fella' could have a pretty good weekend in Vegas with all that stuff.
Good question! We're working on this problem, among other things, at the PSAM project. We have a project to produce high-quality benchmark corpora for spam filter testing. Watch that space for ongoing work, or e-mail us an offer to pitch in and help---we could use it!
When these factors are considered, I think it's quite possible to write software that in the long run has a higher success rate than a human who has better things to do than filter his mail all day.
Dolby noise reduction works by filtering a spectrum into a bunch of bands, each of which are compressed (in an audio sense, not in a digital sense), and recorded to tape. On playback, they go through an expander...how does that concept translate to spam filtering? It can't be "dolby-type", that doesn't make any sense...
ZuluPad, the wiki notepad on crack
What you're planning has already been done, it's called TMDA, and it's not such a good idea. You're going to send out 800 "challenge" emails per day - have you given any thought to how many of those will be genuine addresses, but have nothing to do with the spam you receive because they just happen to be the joe-job victim? These kind of challenge/response systems may slighlty alleviate your own suffering through spam, but at a cost to all those unfortunate enough to have had their email addresses faked. And if the sheer impoliteness of such net behaviour doesn't put you off, note that you're using up more of your own bandwidth to send out such challenges
If any of the smtp exchange or address lookup fails, just forget it, they're probably not real anyway
It would make a lot more sense to make these kind of checks when you're receiving the email in the first place. Reject at the SMTP level - you never accept and process the spam in the first place
My next sig will be ready soon, but subscribers can beat the rush
How would it know if I consider brunettes non-spam but blondes spam? I did opt-in for one of those email categories, but not the other.
Having such a powerful statistical spam filter is definitely a luxury. I have no difficulty believing the accuracy values presented here. I have had experience with spamprobe, CRM114, bogofilter, spambayes, and spamassassin and all of these do an amazing job to the point where spam no longer exists (for you).
Which leads to me plug a little project called WPBL that uses exactly these types of statistical spam filters to spot spam sources in a distributed fashion. Each project member uploads hourly the IPs they see relaying spam and non-spam, where the 'decision' is made by these extremely reliable filters. This effectively converts your regular mail account into an intelligent spam-trap that feeds a central blocklist.
The more members we get, the better we can identify active spam sources around the world. This information is then used by some sites for quite large-scale blocking. Since you're doing all this filtering processing anyway, why not also share "what you learn" (the IPs that are spamming you)?
If this grabs your interest, read up on the reporting scripts or alternatively, the open WPBL data upload protocol if you want to code your own report generator. Bandwidth usage is minimal.
$5 / month hosted VPS on linux = awesome!
Not comparable. The job of a junk mail filter is to drop things I don't want to read. It is trying ot match my evaluation, not to match a semi-objective criterion like red or blue.
If I read 1000 messages and say which I wish I hadn't read, then I am 100% accurate by definition.
Of course, if they are really talking about a pure spam filter -- ie one which identifies unsolicited commercial email -- then they can be more accurate than me, but at an uninteresting, perhaps even counter-productive, task:
I may get unsilicited commercial email I do want to read one day. Almost happened once (I had inadvertantly signed up for it, so it was not really unsolicited, and I didn't actually buy the piece of kit they had on special offer that week, but was tempted). I also get stuff I don't want which isn't spam (notably email from virus infected machines).
The referenced study seems to be a very sloppy job from this POV. They don't define what their criterion of sucess is, and to the extent they put in a hand waving attempt it is clearly nonsense:
`Unsolicited' does not imply `not desired'. If they don't tease those two apart, they can't get interesting results for real world applications. Eg, someone mailing my work address with a commercial proposition may well be a very welcome unsolicited commercial email._O_
.|< The named which can be named is not the true named