taint.org: Justin Mason’s Weblog

• Cycle Helmets and Other Religious Symbols : there appears to be a lack of published research suggesting that bike helmets help avoid serious injury and death — in fact, research seems to suggest the _opposite_.
  (tags: cycling helmets safety research bikes equipment)

• Clever method of near duplicate detection : ‘SIGIR 2008 paper, “SpotSigs: Robust and Efficient Near Duplicate Detection in Large Web Collections”‘. may be useful, although we’ve pretty much stopped deduping in SpamAssassin nowadays
  (tags: corpora dupes duplicates spotsigs collections sigir papers via:jzawodny)

• HadoopStreaming : ‘Using the streaming system you can develop working hadoop jobs with extremely limited knowldge of Java. [..] Hadoop basically becomes a system for making pipes from shell-scripting work (with some fudging) on a cluster.’
  (tags: hadoop perl streams unix distcomp clusters mapreduce)

• political zealot using GMail’s “this is spam” button to deliberately cause spamfilter problems for Obama’s campaign : ‘Tablemate at benihana confided how he subscribes to Obama’s mailing list and marks it all as spam to train Gmail. Urge to kill rising.’ – Kevin Fox on Twitter
  (tags: twitter kevin-fox foaf-story benihana obama anti-spam filtering this-is-spam us-politics moveon)

• “Jake Leg” : ‘large numbers of [adulterated Prohibition-era alcohol, Jamaican Ginger Extract] users began to lose use of their hands and feet. Some victims could walk, but they had no control over the muscles which would normally have enabled them to point their toes upward. Therefore, they would raise their feet high with the toes flopping downward, which would touch the pavement first followed by their heels. The toe first, heel second pattern made a distinctive “tap-click, tap-click” sound as they walked. This very peculiar gait became known as the jake walk and those afflicted were said to have jake leg’
  (tags: jake-leg walking history prohibition alcohol odd bizarre adulteration poison 1930s)

Tags: 1930s, adulteration, alcohol, anti-spam, benihana, bikes, bizarre, clusters, collections, corpora, cycling, distcomp, dupes, duplicates, equipment, filtering, foaf-story, hadoop, helmets, history, jake-leg, kevin-fox, mapreduce, moveon, obama, odd, papers, perl, poison, prohibition, research, safety, sigir, spotsigs, streams, this-is-spam, twitter, unix, us-politics, via:jzawodny, walking

BBC: Cocaine on ‘100% of Irish euros’:

One hundred percent of banknotes in the Republic of Ireland carry traces of cocaine, a new study has found.

Researchers used the latest forensic techniques that would detect even the tiniest fragments to study a batch of 45 used banknotes.

The scientists at Dublin’s City University said they were “surprised by their findings”.

Also at RTE, Irish Examiner, PhysOrg.com, Bloomberg.com, even at Kazakhstan’s KazInform.

This story is (of course) being played widely in the media as “OMG Ireland must use more coke than anywhere else” — in particular, in comparison with a previous study in the US:

The most recent survey carried out in the US showed 65% of dollar notes were contaminated with cocaine.

The DCU press-release has a few more details:

Using a technique involving chromatography/mass spectrometry, a sample of 45 bank notes were analysed to show the level of contamination by cocaine. …

62% of notes were contaminated with levels of cocaine at concentrations greater than 2 nanograms/note, with 5% of the notes showing levels greater than 100 times higher, indicating suspected direct use of the note in either drug dealing or drug inhalation. … The remainder of the notes which showed only ultra-trace quantities of cocaine was most probably the result of contact with other contaminated notes, which could have occurred within bank counting machines or from other contaminated surfaces.

However, looking at an abstract of what I think is the paper in question, Evaluation of monolithic and sub 2 µm particle packed columns for the rapid screening for illicit drugs — application to the determination of drug contamination on Irish euro banknotes, Jonathan Bones, Mirek Macka and Brett Paull, Analyst, 2007, DOI: 10.1039/b615669j, that says:

A study comparing recently available 100 × 3 mm id, 200 × 3 mm id monolithic reversed-phase columns with a 50 × 2.1 mm id, 1.8 µm particle packed reversed-phase columns was carried out to determine the most efficient approach … for the rapid screening of samples for 16 illicit drugs and associated metabolites. … Method performance data showed that the new LC-MS/MS method was significantly more sensitive than previous GC-MS/MS based methods for this application.

My emphasis. I’d guess that that means that comparing this result to banknote-analysis experiments carried out elsewhere using different methods is probably invalid — perhaps this method is more efficient at picking up ‘contact with other contaminated notes, which could have occurred within bank counting machines or from other contaminated surfaces’, as noted in the DCU release?

Tags: bad-journalism, banknotes, chemistry, cocaine, dcu, drugs, ireland, press, research, science

Science: in passing — this came up elsewhere, and it’s worth copying here, too (for reference).

The question was: how much should publicly-funded researchers be required to disclose – should they be allowed to generate ‘closed-source’ solutions at the taxpayers’ expense?

In the US and world-wide, there used to be a tradition that government-funded research should be made open to all, since if it was funded from public taxation, the fruits of that taxation should go back to the public. However, 25 years ago, the US enacted the Bayh-Dole Act, in which:

• Universities were encouraged to collaborate commercial concerns to promote the utilization of inventions arising from federal funding.
• It was clearly stated that universities may elect to retain title to inventions developed through government funding.
• Universities must file patents on inventions they elect to own.

So in other words, the government has dictated since 1980 that government-funded research should not produce open-source or public-domain solutions, necessarily, as the results of research are to be considered private-sector profit-generating centers for the host universities. Naturally, cash-strapped universities have imposed internal regulations to maximise revenue from their research staff.

The implications for whatever ‘the next BSD TCP/IP stack’ may be are obvious.

Tags: copying, expense, funding, government, public, question, reference, research, science, taxation, tradition

Games: OK, JFK Reloaded is very, very wierd.

Read the insanely detailed FAQ and boggle at the author’s obsessive research and fetishistic recreation of the events at Dealey Plaza, November 22nd 1963.

Quite worrying, to be honest!

Tags: author, boggle, dealey, faq, games, jfk, plaza, recreation, reloaded, research, wierd

Spam: So, CEAS was great fun, and very educational:

• Got to meet up with various antispammers, including Daniel and Theo from the SpamAssassin dev team, Jeff Chan from SURBL, Dan Kohn from Habeas, Catherine Hampton from The SpamBouncer, Miles Libbey, John Levine, Neil Schwartzman — lots of good chats.
• MS really know how to feed a conference! I hear rumours there was an extra-special tinned-meat-product-based dish at the banquet…
• But their firewalling tendencies put a serious damper on keeping in touch with the outside world, at least until we set up an SSH tunnel on port 443 ;)
• During a lull, Dan Kohn fired off a hands-up census — a good 75% of the attendees (roughly) admitted to using SpamAssassin!

My highlight papers:

• IBM’s Chung-Kwei pattern-discovery system — the one which Mark dug up. Very interesting stuff; it turns out that bioinformatics is full of large corpora of data (genomes) which you then need to find patterns in. Funnily enough, so is SpamAssassin: s/genomes/spam/, s/patterns/regular expressions/. The more advanced pattern-discovery algorithms even allow complex patterns to contain alternative blocks, ‘don’t-cares’ and similar regular-expression-like features.

  The really good bit of Chung-Kwei is the Teiresias algorithm (more pages, online demo). Of course, being IBM research, it’s probably patented to the hilt, and may be tricky to license; but it’s certainly pointed us in a whole new interesting direction — anyone know any bioinformaticians?

  IBM is really gearing up on anti-spam research. 4 of the 6 papers listed on their website were presented this year, at CEAS.

• Another good paper was On Attacking Statistical Spam Filters, by Gregory L. Wittel and S. Felix Wu, which (similarly to Henry Stern’s submission, which I helped a little with) dealt with an attack on Bayesian filters.

  This is interesting stuff; we’re pretty sure it’s not as serious as it could possibly be, in SpamAssassin’s implementation, but it’s still a serious attack.

• The Impact of Feature Selection on Signature-Driven Spam Detection was an interesting paper on AOL’s new signature schemes. (The conference was sponsored by Cloudmark, BTW, but those guys were nowhere to be seen — in which case they missed this presentation ;)
• Reputation Network Analysis for Email Filtering was interesting, in that it mirrors to a degree the thinking behind web-o-trust.org, but in my opinion suffered due to a lack of thought about avoiding spoofing (by including IP address information in the FOAF file, it could do this now). However, once SPF becomes pervasive, this could be combined with that to generate personalised webs of trust usable for email whitelisting.
• Resisting SPAM Delivery by TCP Damping was very nifty; plug a classifier into your MTA, and thereby detect connections from spam relays. Once you’ve found them, you then throttle down their connection as they attempt to deliver spam. Some other TCP-level tricks can do nifty stuff like massively increasing the bandwidth consumption of the spamming machines. Very very nice!

I took copious notes on the SpamAssassin wiki, if anyone’s curious.

Tags: attack, ceas, dev, fun, got, ibm, research, spam, spamassassin, stuff

Spam: Some fantastic data in this paper from the Kentucky Long-Term Policy Research Center.

It’s a brief 2-pager detailing the effectiveness of the CAN-SPAM Act in reducing the spam load, using a set of test addresses. The methodology is pretty good.

One point in particular is very important: ‘opting out’ from spam Just Does Not Work. This graph tells the whole story:

After opting out from spams received, the amount of spam received at those ‘opted out’ test addresses actually rose. (This even after CAN-SPAM made such activity explicitly illegal.)

Some other data:

• obfuscating addresses on web pages is still working; 7.7 times the spam is received if you don’t bother doing so.
• e-mail harvesting also continues after CAN-SPAM made it illegal.

If anyone needed proof, this shows that spammers are quite happy to break the law; strong enforcement ‘teeth’ are needed for any anti-spam legislation. (UK, take note: the thoroughly useless system whereby spam complaints must be submitted on paper isn’t going to help!)

The Technical Details document also notes something interesting: one test address was set up to test ‘opting out’ of legitimate mass mail from some (unnamed) big websites, and continued to receive ads ’sometimes months after opting-out’. For shame!

(thx to John Levine for forwarding the links.)

Spam: Michael Radwin on open HTTP redirectors, and in particular noting that Yahoo! have (finally) closed their main one down. One down, several hundred to go ;)

Good history of the exploitation techniques that spammers have been using, too.

Tags: can-spam, data, kentucky, long, paper, policy, research, spam, term, test

Funny: AP: SoCal city falls victim to Internet hoax, considers banning items made
with water. It’s the old ‘dihydrogen monoxide’ hoax again:

‘It’s embarrassing,’ said City Manager David J. Norman. ‘We had a paralegal who did bad research.’

The paralegal apparently fell victim to one of the many official looking Web sites that have been put up by pranksters to describe dihydrogen monoxide as ‘an odorless, tasteless chemical’ that can be deadly if accidentally inhaled.

So — ha ha, stupid Aliso Viejo city officials. But seriously — why is a paralegal making decisions on scientific issues? Isn’t that what the EPA and their environmental scientists are there for? Tail wagging the dog, I think.

Tags: city, dihydrogen, funny, hoax, internet, monoxide, paralegal, research, socal, victim, water

Tech: ATAC: Abusable Technologies Awareness Center. Great panel weblog, with some of the big names in the research field, dealing with several security issues quite nicely.

Found from a link to Simon Byers’ 2003 roundup of information leakage, which notes an interesting case I hadn’t heard of – a TechTV presenter accidentally posting topless photos of herself, due to a bug in Photoshop!

(link via Liudvikas Bukys)

Tags: abusable, atac, awareness, center, field, link, panel, research, security, tech, weblog

Beer: Irish Independent: Now ads can’t say it but you always knew it — Guinness IS
good for you:

One pint of Guinness a day can reduce the risk of blood clots that cause heart attacks, according to new research presented at the annual meeting of the American Heart Association in Orlando, Florida.

… Scientists investigating the health benefits of drinking beer found that stouts like Guinness worked much better than lager. They said dark beers were packed with anti-oxidant compounds called flavonoids which help reduce damage to the lining of the arteries. … For maximum benefit a person would need to drink just over one pint of Guinness a day.

My grandfather was ‘prescribed’ a bottle of Guinness per day by his GP, to lower cholesterol and blood pressure. Mind you, that was in ’70s Ireland ;)

Tags: beer, blood, day, guinness, heart, independent, irish, meeting, pint, research, risk

Amazing! Porton Down is the UK’s center for research into chemical and biological weapons, and has been since 1916. Not the nicest place you could think of — by a long shot.

Well, it turns out that the massive no-go buffer zone around Porton Down, existing for 87 years, has preserved ‘the largest remaining continuous tract of chalk downland in Britain’. ‘The farming revolution of the 20th century, the development, the tourism, have all passed it by.’ ‘The disrupters are the large-scale inputs of chemicals, the pesticides, herbicides and artificial fertilisers that are the essence of intensive farming. At Porton Down, these have never arrived.’

As a result, it’s now an amazing wildlife heritage site. Quite hard to get to see it — but good to know it’s there! Thanks to Bruce Sterling for forwarding this along the Viridian list.

Reminds me of something I heard about Chernobyl — since the area around it is heavily irradiated, and therefore a no-go area for humans, it’s become a de-facto wildlife refuge (even if half of the animal inhabitants are sterile as a result.)

Tags: area, center, chemical, farming, nicest, place, porton, research, result, wildlife

The Dublin Flash Mob. All went off very well, from the sounds of it. However, this picture contains some wierdness — who the hell is that guy, second from the left, who’s stolen my haircut circa 2 years ago?! Those are my sideburns, give ‘em back!

(ObSoCalJoke: they tried to organise a flash mob in southern CA, but couldn’t find anywhere with a big enough parking lot for all those single-occupant SUVs. Ba-dum-tish!)

Telecoms: The Communications Workers of America union have released some figures on Verizon’s profit margins etc. Interesting to note some figures — like they charge 4 dollars for call waiting, a service which costs them 0.82 of a cent to provide — that works out at a 48,680% profit margin, which must be nice. In addition, Verizon use ’splitters’, which result in a copper pair being unusable for DSL — just like Eircom do in rural Ireland. Interesting to note that, even after deregulation, LLU and general introduction of competition, the same problems still arise.

Science: BBC: Scientific research put under spotlight. Terrible article from the Beeb, who should know better.

Basically the article pins some of the blame for recent absurd claims of scientific breakthroughs, like the Raelian’s claims they cloned a human, on the peer review process.

What they’re missing is that, in most cases of these absurd claims, the research had not been peer reviewed — instead a press release was put out in advance. Peer review remains the most effective way to demolish bad science. However, the news media shows no sign of being willing to sit around and wait for other scientists to analyse the latest claims, before publishing them.

Spam: Salon: Meet The Spam Nazi. More on the bizarre story of the Jewish leader of a Nazi party, who now peddles ‘make penis fast’ pills.

Politics: Ian ‘Freenet’ Clarke says he’s leaving the US.

Linux: I’ve given up on blogging the SCO-v-everyone thing, it’s getting too absurd. GrokLaw is covering it much better than I could anyway. Plus: You say po-TAY-to, I say po-TAH-to.

Movies: I concur with Waider — Pirates of the Caribbean is great. Best summer blockbuster in years; Hollywood can still pull off a good big movie now and again (by using young directors it seems). Buckle those swashes! Aarrr!

Tags: article, flash, mob, peer, profit, research, review, science, spam, verizon

German euros ‘full of cocaine’ (BBC):

Almost all euro banknotes circulating in Germany contain traces of cocaine, German researchers say. … ‘Nine out of 10 banknotes show clearly measurable amounts of cocaine,’ Professor Fritz Soergel of the Institute for Biomedical and Pharmaceutical Research in Nuremberg was quoted as saying by Reuters news agency.

… The concentrations of cocaine on Spanish euro notes were almost a hundred times that of what was recorded in Germany; … Professor Soergel said that his team was ‘almost knocked flat’ by results of yet another recent study in Barcelona.

Tags: agency, bbc, biomedical, cocaine, euro, institute, news, pharmaceutical, research, reuters

Forbes: A Patent On Porn. It seems Acacia Research, an intellectual-property ’shell’ company, has a bunch of crappy software patents on streaming media (to go with their patent on the ‘V-Chip’, remember that?).

Things haven’t been going too good recently. Apparently, they decided to ‘monetize’ these streaming-media patents — in other words get all Sopranos on a bunch of small players, namely 700 porn site operators, sending some legal threats to ‘pay up — 1-2% of gross — or get sued’ their way.

What happened? Did the pr0nsters roll over and cough up? Not a hope.

Eight firms (of 700) agreed to Acacia’s terms. But 40 didn’t, and Acacia promptly slapped them with lawsuits. Rather than buckling, though, several of the porno sites joined together and stood their ground. Now Acacia is in the fight of its life and may even face a shareholder revolt as a result.

Read on for the rest…

Tags: bunch, company, crappy, forbes, patent, porn, research, shell, software, the

and now, on a lighter note, The Observer reports that the ‘cannabis economy’ in the UK is worth 11 billion UKP a year:

A major new study is being used to advise well known household and high-street companies about the gains and losses they face as cannabis smoking becomes commonplace. Research has revealed that Britain’s ‘cannabis economy’ is worth 5 billion a year in sales alone. Now it has been discovered that a further 6bn of consumer expenditure each year is closely linked to the growing cannabis-users’ market.

‘Young people between 15 and 30 are very trend-conscious and aspirational,’ said Andy Davidson, who commissioned the study for The Research Business International, trend analysts who tracked the spending habits of young people for six months.The study found that cannabis users spend an average of UKP 20 on products that accompany their drug use each time they smoke.

Because smoking cannabis heightens appetite, users are providing a UKP 120 million weekly windfall to a string of takeaway food suppliers, such as Domino and Pizza Hut, and manufacturers of ‘munchie’ products such as Mars bars and Haribo jellies.

Tags: economy, household, note, observer, research, smoking, study, the, ukp, year

‘Spam’ Likely to Clutter E-Mail for Some Time, says Jupiter Research (via Reuters).

“It’s getting easier to send spam messages. You can buy a CD-ROM with millions of e-mail addresses for next to nothing and send it out for next to nothing,” said Jared Blank, senior analyst at Jupiter.

“Spammers are clever people and there is clearly an arms race between spammers and people trying to prevent spam that just constantly escalates,” said Forrester analyst Jim Nail. “Having simple lists of spammers and domains — that’s not enough because spammers change domains or addresses to stay ahead.”

So, good news: I have a job. Bad news: well, I think that side is obvious ;)

Tags: ail, analyst, clutter, jupiter, news, nothing, research, reuters, spam, time

BB reports that “Russian entrepreneurs are spraypainting logoed advertisements for their products and services on stray dogs and releasing them as walking, starving billboards.” This sounds just a bit too Chris Morris to me, and considering it came via Ananova / Orange Today’s “quirkies” service – which is not exactly reknowned for doing the backup research first – I would say it’s pretty unlikely… let’s see what forteana make of it.

Tags: ananova, backup, bit, forteana, make, orange, research, service, today

Geek hero:

The publication in Genome Research gives details of (Jim) Kent’s algorithm as a demonstration of openness, which has been a hallmark of the public Human Genome Project.

“Instead of being a black box it details how it was done,” said John McPherson, co-director of the genome sequencing center at Washington University in St. Louis, Mo., one of the many labs that contributed to the Human Genome Project.

The free exchange of information is a testament to why Kent became passionate about the public Human Genome Project in the first place.

“I thought it would help to get as much information about genes and the genome in to the public domain to help discourage people from patenting it wholesale,” Kent said.

“I was afraid that if the only people who had access were the people who could afford Celera’s (subscription) database, it would tie things up.”

Sorry, it’s old bits, but I hadn’t read it before.

Tags: algorithm, demonstration, geek, genome, hero, human, information, project, publication, research