Share This

Thursday, 18 February 2010

Scientist Finds PageRank-Type Algorithm from the 1940s

Google's PageRank algorithm was developed in 1998. But a project to trace the history of such algorithms reveals an example from the 1940s.
The PageRank algorithm is a key part of Google's method of ranking web pages in search results. It uses the network of links between web pages to determine their value and, famously, judges a page to be important if it is linked to by other important pages.

One crucial feature of this idea is that it requires an iterative approach to constantly re-evaluate the value of a page as the importance of others varies. Iterative ranking algorithms have since become an important part of network theory.

PageRank was developed in 1998 by Google's founders Sergey Brin and Larry Page and its impact has been such that it's easy to forget that the approach was not entirely novel. Massimo Franceschet at the University of Udine in Italy points out that the idea has been successfully exploited a number of times in 20th century science, even before Brin and Page were born. Today, he presents a short history of iterative ranking algorithms and charts their evolution prior to Google's emergence.

He begins in reverse chronological order with the work of Jon Kleinberg, a computer scientist at Cornell University, who developed an almost identical approach to PageRank, just a few years earlier. Brin and Page even reference his work in their famous paper introducing PageRank.

Kleinberg called his algorithm Hypertext Induced Topic Search or HITS and it treated web pages as "hubs" and "authorities". It used the circular definition that authorities are pages that are pointed to by hubs and hubs are pages that point to authorities and requires an iterative approach to solve.

In the heady days of the dotcom boom in the late 20th century, before Google became so successful, Kleinberg's work received considerable media coverage.

Franceschet also examines the work of Gabriel Pinski and Francis Narin who developed a way of ranking journals. Their rule was that a journal is important if it is cited by other important journals. Like PageRank and HITS, this requires an iterative method to exploit the structure of links between journals to come up with a ranking.

Long before this, however, Charles H Hubbell at the University of Califronia , Santa Barbara, was analysing social networks in a similar way. In 1965, he published a technique for determining the importance of individuals based on the importance of the people who endorse them. This again has the characteristic circular definition and iterative solution. Hubbell is acknowledged by many including Kleinberg as a pioneer in iterative ranking theory.

But the big surprise is Franceschet's discovery of an even earlier forerunner to PageRank in the work of the Harvard economist Wassily Leontief. In 1941, Leontief published a paper in which he divides a country's economy into sectors that both supply and receive resources from each other, although not in equal measure. One important question is: what is the value of each sector when they are so tightly integrated? Leontief's answer was to develop an iterative method of valuing each sector based on the importance of the sectors that supply it. Sound familiar? In 1973, Leontief was awarded the Nobel Prize in economics for this work.

What's clear is that the ideas behind PageRank have a venerable history but the surprise is that they date back to at least the 1940s. It'll be interesting to see if anybody can find any similar work that predates this.

Ref:arxiv.org/abs/1002.2858: PageRank: Stand On The Shoulders Of Giants

Tuesday, 16 February 2010

Symbian 3 Smartphone Platform Released at Mobile World Congress

RIM Introduces BlackBerry Enterprise Server Express


King of Cheez: The Internet’s Meme Maestro Turns Junk Into Gold

Photo: Misha Gravenor

What most people view as a workday time-suck, Ben Huh sees as a potential gold mine.
Photo: Misha Gravenor

Just how funny is a beer-drinking horse? Ben Huh is sitting in his downtown Seattle office asking himself this question. It’s an unseasonably warm afternoon in November, and Huh, the 32-year-old founder of the humor-blog startup Cheezburger Network, is deciding whether a picture of a boozy equine chugging a cold one should run on Daily Squee, a Web site devoted entirely to user-generated snapshots of twee creatures.

The problem with drunken farm animals, though, is that they’re never quite as cute as you’d hope. “This is kind of close to animal abuse,” Huh says, pivoting in his chair in mild disgust. He’s dressed in dark blue jeans, cream-colored Warhol-replica eyeglasses, and a red T-shirt featuring a freakishly long-torsoed kitty — a nod to lolcats, one of the many Web phenomena Huh has made mainstream.
He turns to Kiki Kane, who manages new site development at Cheezburger. “A horse drinking beer is not Daily Squee,” Huh determines.

He and Kane then half-jokingly brainstorm some possible new sites that might run this somewhat unsavory image: WTFnature.com? Naturedoingitsownthing.com? The conversation quickly moves on to the recent influx of user-generated dog-humping pictures. “People submit 500 pictures of dogs humping every day!” Kane says. “There’s got to be a place for those.”

“There is,” Huh replies. “But it’s not with us.” As he later explains: “We’ve done this enough times to know that’s just a one-note joke.”

Source: Nielsen
Source: Nielsen

For almost three years now, this has been Huh’s life: to pore over millions of JPEGs and YouTube clips in search of Internet memes — those absurd running gags that hatch and proliferate on the Web seemingly overnight — and figure out which of these quick-hit laughs might yield long-term profits. Since it launched, the Cheezburger Network has successfully aggregated more than 30 sites. You’ve likely visited a few of them, perhaps at the behest of an easily distracted coworker or a walrus-loving aunt. There’s Huh’s flagship site, I Can Has Cheezburger?, a vast repository of lolcat images (for the uninitiated, these are cat pictures with absurd, syntactically challenged captions). There’s GraphJam, a data-visualization blog that renders witty pop-culture musings into pie charts, Venn diagrams, and illustrated maps. The aptly named FAIL Blog — the Cheezburger Network’s most popular site, with 1.1 million visitors per month in the US — runs a seemingly infinite number of skateboard spills, nut-smacks, and hilariously misspelled signs.

What most people view as a workday time-suck, Huh sees as a potential gold mine. And so far, he’s been right. Back in September 2007, when most tech players were pouring VC money into the next Facebook, Huh hooked up with a group of angel investors to buy his first fledgling phenomenon, the I Can Has Cheezburger? blog. Since then, he has built Cheezburger Network into the largest aggregator of Web memes, pulling in more than 200 million pageviews a month combined. Huh won’t divulge financial specifics, but investor documents show the Cheezburger Network approached $4 million in revenue last year. The money comes from display ads (companies like American Express and Burger King sponsor the sites) as well as books (the lolcat series has produced two New York Times best-sellers), T-shirts, and other merch. “In the past year and a half, a meme industry has come into place,” says Tim Hwang, organizer of ROFLcon, a biannual Web-celebrity gathering. “There are people who are interested in commercializing memes. Ben’s accelerating that development. Suddenly, everybody’s like, ‘Wow, you can actually pursue this for a living.’”

Of course, Huh can’t take all of the credit for Cheezburger’s success. In fact, he owes quite a bit to the millions of anonymous Web dwellers whose work he corrals, curates, and posts. The majority of Cheezburger’s sites are, after all, extensions of ideas born on ungoverned image-board sites like 4chan or Something Awful — inside jokes that bubble up, JPEG by JPEG, into the mainstream. Lolcats, for example, are an offshoot of Caturday, a 4chan chestnut that dates back to at least 2005. FAIL is a long-running Web gag traceable to Blazing Star, a 1998 Japanese videogame that taunts players with onscreen messages like “You fail it!” Other sites, like GraphJam or the subversive motivational posters of Very Demotivational, were rough concepts until Huh figured out how to develop and package them.

Huh’s setup encourages users to submit their own lolcats or FAIL entries, ensuring a continuous supply of content. “I used to want to create memes more,” he says. “But what’s more satisfying: playing on the playground or building a playground for a bunch of people to enjoy? I’m much more the person who’d rather build it. That brings me satisfaction.”


The playground metaphor is apt: The Internet is supposed to be a great place for sharing and disseminating, which is how a joke evolves into a meme in the first place. Huh has actually been accused of being a bit of a schoolyard mooch — sponging up clever ideas that don’t belong to him (or anyone, really) and dispatching them to mainstream (read: lame) audiences for his own personal gain.

Huh doesn’t take credit for inventing lolcats or any of the other trends he has adopted and adapted. Nonetheless, he has numerous online critics who have expressed their displeasure by subjecting Huh and his wife to flame wars, denial-of-service attacks, and death threats.

Photos: Misha Gravenor
Scenes from the office, clockwise from top left: scheduling a post on thatwillbuffout.com; T-shirt designs for LOLmart Shirts; sticky note idea board; brainstorming whiteboard.
Photos: Misha Gravenor

Most of this has been organized, not surprisingly, by various so-called /b/tards on 4chan’s /b/ board, a sort of online Mos Eisley where members of the Anonymous griefer movement have previously planned raids against other ideological foes, like Scientology and YouTube. To many of them, Huh is a poseur who is exploiting and overexposing their underground culture. “When someone throws their name on something that’s been around for a while and that’s not really theirs, it pisses off the people who liked it when it was a more pure thing,” says a longtime 4chan participant who goes by the name Sethdood. “Anonymous doesn’t want to be associated with anything that cleans itself up for kids or that’s goofy and nice. They’ll disown it as soon as it doesn’t become theirs anymore.” Another veteran 4channer, who goes by the handle Kakama, compares Huh’s company to a teenybopper mall franchise: “Huh’s enterprise is the Hot Topic of the Internet. Every time we walk around and hear some random guy going ‘LOL! I can has cheezburger!’ it’s disgusting. It’s like a little bit of our culture has been taken out and defiled.”

Huh refuses to let the haters get to him. “This is all part of the game,” he says. “And if I were scared, what am I going to do? I mean, it’s the Internet. If somebody wants to come find me, somebody will.”

The Cheezburger Network headquarters is located on the second floor of a five-story office building. This afternoon the crew of 10 Cheezburger moderators — all in their early to mid-twenties — are clustered in one corner, headphones slung around their necks. The moderators spend hours drilling into the inner core of YouTube and wading through the thousands of user-uploaded photos and videos on Cheezburger’s various sites in search of one amazing thing to snap up. Every few minutes, someone cues up a warbled home-video musical performance or old videogame theme song. Huh also has 10 writers scattered around the country, some of whom are plucked directly from the meme world (Brad O’Farrell, originator of the “Keyboard Cat” meme, oversees Daily Squee).

“What are the rules for butts on This Is Photobomb?” asks a tall, bespectacled moderator named Steve Ibsen, referring to a site that collects crude and candid party pictures.

“If there’s a crack,” Kane replies, “you gotta cover it up.”

Huh is sitting just outside the windowless former server closet he now calls his office. It’s not the most august perch for a CEO, but years ago Huh learned the pitfalls of executive excess. In January 2000, after graduating from Northwestern University with a journalism degree, he founded his own analytics startup. When the Nasdaq crashed that spring, Huh was forced to close up shop. “It was an abysmal failure,” he says. “I hired too many people. I didn’t raise enough money. We didn’t actually have a product.”

Huh spent the next several years moving around the tech industry, from an Internet-radio startup to a software-installation firm. In early 2007, he and his wife started a modest pet-news blog called Itchmo. The site became a must-read for animal lovers, including Eric Nakagawa, one of the original cofounders of I Can Has Cheezburger?

Nakagawa and his partner, Kari Unebasami, had launched Cheezburger in 2007 after spotting a Something Awful image of a wide-eyed, overly excited kitty accompanied by the caption “I can has cheezburger?” They began collecting other lolcat pictures from sites like 4chan and Something Awful and installed their own lolcat builder for visitors to slap cat patois captions on their own photos.

When Nakagawa linked to an Itchmo post in May, Huh’s site was so flooded with traffic that it crashed. After corresponding with Nakagawa and Unebasami, Huh learned that Cheezburger’s traffic was exploding — and the owners were overwhelmed. “At that point, I was putting in 20 hours a day and not getting a lot of sleep,” Nakagawa says. “We were getting a few thousand pictures a day. We were a little burned out.”

Sensing that lolcats had crossed from radar-blip Web fad to full-on phenomenon, Huh decided to seize the opportunity. In August he made Nakagawa an offer over IM to buy I Can Has Cheezburger? (and its sister site called I Has a Hotdog!). He put down $10,000 of his own money, got some investors, and the deal was finalized a month later. The Cheezburger Network was born. Neither Nakagawa nor Huh would disclose the financial details, but published reports put the purchase price as high as $2 million.

By February 2008, the Cheezburger Network had launched GraphJam and Pundit Kitchen, where users could insert droll commentary onto snapshots of candidates Sarah Palin and Barack Obama. Meanwhile, Huh scoured the Internet for more material he could transform into memes. “I thought, ‘Dude, there’s so much more of this stuff — why aren’t we doing it?’” he says.

Photos: Misha Gravenor
Cheezburger moderators Joe Olk and Lisa Kacerosky search YouTube and thousands of user-upload photos and videos for material.
Photos: Misha Gravenor

As Cheezburger expanded, Huh refined the company’s modus operandi: Rip off a concept from the Web and seed a site devoted to the idea with material that’s already floating around. Next, if needed, sharpen the content with snark produced by a stable of moderators and writers. Then turn it over to users to riff on the meme. Huh has attempted about 40 sites so far, not all of them successful: The Twitter-wiki parody 140pedia, for example, flopped. Huh says he’ll give a floundering site a month or two before shutting it down and repurposing the content.

Right now, the Cheezburger Network pretty much pwns the meme-aggregation marketplace, but it’s a wobbly dominion. Theoretically, anyone can set up a Tumblr account and, with an hour or so of Web surfing, create a clearinghouse for Hitler Downfall parodies. Memes can now spread more rapidly (and therefore risk flaming out just as quickly) thanks to real-time tracking sites like the Daily Meme and Know Your Meme. Which means that for Huh, the real challenge is not in figuring out what’s funny but sussing out the exact moment something will jump from image-board fringe to moms-forwarding-it ubiquity.

The tactics that Huh uses to beat out the other meme-jockeying Web sites can be a little obnoxious. Consider the case of Engrish.com. Last year, Huh tried to acquire the long-running botched-translation site, but when he and the owner couldn’t agree to terms, Huh simply set up his own site, Engrish Funny. “We’ll launch a Web site and people will be like, ‘You’re copying these guys!’” Huh says. “But they’re copying these guys, and the guys before that were copying these guys. Everything we do is some variation on the past.”

Huh has a point. The Cheezburger sites are recycling decades-old comedic constructs. The celebrity-doppelgänger site Totally Looks Like is a riff on Spy magazine’s Separated at Birth? franchise. FAIL Blog, meanwhile, has co-opted America’s Funniest Home Videos and National Lampoon’s True Facts section.

Huh has about 150 other ideas in development and about 1,000 registered domain names. He has also been talking with Hollywood producers about expanding his Cheezburger brands into television series. Nowadays, people don’t forward memes to Huh; they pitch them. Usually, he preempts them by saying that he’s probably already heard it before — and that he’s already working on it himself. “If you see a similar site in the future, it doesn’t mean we took your idea,” he says. “And if you’re OK with that, then you can tell me.” Just don’t try boinking-pooches.com. That’s a guaranteed FAIL.

Source: Contributing editor Brian Raftery (brian raftery@gmail.com) wrote about B-movie production house Asylum in issue 18.01.




EU's Financial Woes

EU's Financial Woes 

Sovereign debt crisis to derail world growth?

THE sovereign debt crisis contagion is spreading in Southern Europe (see charts), from Greece to Portugal, Spain and Italy, where government debts and budget deficits are high.

Investors have sold government bonds in those countries as perceived default risks have risen.

This has resulted in the rise in the yields of government bonds resulting in higher borrowing costs for the government and private sector as loans are often tied to the risk free rate of government bonds.

Countries that faced sovereign debt crisis earlier, like Iceland, Ireland, Hungary and Latvia, had to reduce their budget deficits by raising taxes and cutting government spending, resulting in economies going into recessions.
For example, austerity measures in Ireland have resulted in the economy shrinking by almost 12% in the last two years.

Unfortunately, those countries in the eurozone cannot print money like the US, UK and Japan to finance their deficits through the monetisation of debt as they have given up control of their monetary policies to the European Central Bank.

Leaving the eurozone and devaluing their currencies will likely lead to an Argentinian-style loss in confidence and massive fund outflows.

Large budget deficits were the cause of the problems faced by these European countries but still they are likely to resist ceding control over matters like government spending and taxation to the European Union authorities.

As the European monetary union is facing tremendous stress, can the currency union survive where there is no fiscal union or discipline?


Austerity measures are being forced on these Southern European countries at a time when economic conditions are terrible.

For example, following the collapse of the Spanish property bubble which decimated the construction and real estate industry, Spanish unemployment has exceeded four million, representing an unemployment rate of close to 20%.

Cutting Spanish government spending to reduce its 11.4% budget deficit as a percentage of gross domestic product can only worsen unemployment.

The downgrading of the sovereign debts of Southern European countries is also fuelling the contagion and capital flight.

The sovereign rating of Japan, with an extremely high debt to GDP of close to 200%, is at risk with S&P placing a negative outlook on Japan’s double A sovereign credit rating.

Strangely, these agencies have continued to maintain the triple A ratings of the US and UK, even though their budget deficits are very high, estimated at US$1.6 trillion of 10.6% of GDP for the US.

This is despite the fact that the US unlike Japan cannot rely on domestic savings to fund its borrowings. But then, these were the same agencies that maintained AIG’s triple A rating even when it was guaranteeing large amounts of risky subprime loans.

Even the US is facing constraints on the size of its budget deficits as a popular revolt against large deficits are being organised and the loss of the 60-vote Democratic majority in the Senate means that Republicans can now block spending bills.

The world is depending on fiscal stimulus to sustain growth as highly geared Western consumers are still deleveraging and are reluctant to borrow more despite low interest rates.

If enough countries withdraw fiscal stimulus because of the sovereign debt contagion, world growth will stall, plunging the world into a double dip recession.

This possibility cannot be fully discounted as the GDP of all the Southern and Eastern European countries combined is estimated by the International Monetary Fund (IMF) at US$9.1 trillion in 2008, more than twice the size of the Chinese economy.

Already, it would appear that government spending cuts and loss in confidence in the euro has permanently damaged European growth for the next two years.

The EU’s economy, with an estimated GDP of US$18.4 trillion in 2008 is larger than the US economy at US$14.4 trillion.

US and Chinese manufacturing will also be adversely impacted as demand from Europe will decline and a weaker Euro makes US and Chinese exports less competitive.

To prevent this contagion from spreading, the European central bank would have to guarantee the sovereign debt of these affected countries.

Excessive speculation should be curbed to prevent a downward spiral like what happened during the Asian crisis.

Speculative attacks during the Asian crisis ended when speculators who short sold were killed off by the superior buying power of the Chinese and Hong Kong authorities which also changed rules on short selling.

Presumably, Germany and France with the support of other major powers (US, China and the UK) could do the same. That of course would be premised on Greece cutting its budget deficit; probably a foregone conclusion as funding problems will curb spending.

Sounds familiar? The similarities to the Asian financial crisis are as follows: austerity measures were imposed when the economy was hurting and the flight of short-term funds resulted in higher interest rates.

Like Tun Mahathir Mohamad, the Greek prime minister is already blaming unscrupulous hedge funds and speculators for its predicament.

A withdrawal of fiscal stimulus would slow world growth and increase the chance of a double dip recession.

Corporate earnings and the stock market are likely to be adversely impacted. On the flipside, a weak economy will allow interest rates to remain low for a longer time.

The inability to fund fiscal stimulus through government debt may also increase the temptation to print money to finance the deficits, especially if deflationary pressures arise from a slower growth (exacerbated by lower commodity prices and excess capacity).

If such a downturn is moderate, it could lead to better fiscal discipline and a more rapid private sector adjustment; Asian countries hit by the Asian financial crisis (including Malaysia) have weathered the current downturn better as their financial systems are more stable now, budget deficits can be funded by domestic savings, asset bubbles have been better contained and borrowings are funded mainly by domestic sources.

In an environment of weak growth and low interest rates, stick to defensive sectors with steady demand like healthcare (rubber glove companies), utilities, tobacco and telecommunications.

A high dividend yield should also be welcomed as interest rates are likely to remain low due to slower growth and deflationary pressure.

Source: Starbiz ● Choong Khuat Hock, head of research at Kumpulan Sentiasa Cemerlang Sdn Bhd.