Strata Week: 4.74 degrees of Kevin Bacon

Facebook says we're closer than we thought, Gnip targets finance, and eBay grabs Hunch.

Here are some of the data stories that caught my attention this week:

There are less than six degrees between you and Kevin Bacon

You know the game: there are no more than six degrees of separation between actor Kevin Bacon and anyone working in Hollywood. You can start with Sir Alec Guinness, or you can start with Sasha Grey — there are, at most, six links that connect that performer to Kevin Bacon. The game is built on an older notion of social connections, one dating back as far as the late 1920s when Hungarian author Frigyes Karinthy argued there were no more than five acquaintances that separated people.

It might not be all that surprising that the Internet affords different — closer? — relationships than those of the earliest 20th century Hungary. Indeed, according to Facebook’s data team, the connections it now affords are actually much closer than the “six degrees of separation” maxim. Between Facebook relationships, there are only 4.74 degrees (“hops” it calls them).

Facebook’s Data team writes:

Thus, when considering even the most distant Facebook user in the Siberian tundra or the Peruvian rainforest, a friend of your friend probably knows a friend of their friend. When we limit our analysis to a single country, be it the US, Sweden, Italy, or any other, we find that the world gets even smaller, and most pairs of people are only separated by three degrees (four hops). It is important to note that while [Stanley] Milgram was motivated by the same question (how many individuals separate any two people), these numbers are not directly comparable; his subjects only had limited knowledge of the social network, while we have a nearly complete representation of the entire thing. Our measurements essentially describe the shortest possible routes that his subjects could have found.

The Facebook study involves a sizable dataset — some 721 million Facebook users. That’s the largest, by far, of any study of its kind. But as The New York Times points out, the Facebook study still raises questions about what exactly do we mean when we talk about “friends” and the relationships and connections we’ve created between people online.

Strata 2012 — The 2012 Strata Conference, being held Feb. 28-March 1 in Santa Clara, Calif., will offer three full days of hands-on data training and information-rich sessions. Strata brings together the people, tools, and technologies you need to make data work.

Save 20% on registration with the code RADAR20

Stock market data meets social media data

GnipSocial media data aggregator Gnip announced this week that it was expanding into the financial services market with the launch of a new product aimed at hedge funds and stock traders. Gnip MarketStream will provide real-time data from both Twitter and StockTwits.

Gnip is one of two companies licensed to handle the Twitter firehose (the other, which I wrote about last week, is DataSift). The company says that it already includes in its customer base a number of hedge funds, and now in partnership with StockTwits — a financial data platform that created the $(TICKER) tag on Twitter — Gnip says it will provide real-time social data to trading companies.

Hunch acquired by eBay

HunchEBay has acquired the recommendation engine Hunch. The price tag was around $80 million, according to Mike Arrington. Founded in 2007 by Chris Dixon and Caterina Fake, Hunch built a “taste graph,” a website and API that provided insights into users’ affinities with different people, services, brands, and websites.

In an interview with Betabeat, Dixon pointed to precisely this sort of insight as the rationale for the startup’s acquisition: “eBay is a very unique retailer,” Dixon told Betabeat. “When grandma posts a sweater for sale, it doesn’t have metadata to help sort and identify it. In working to understand users’ tastes on the open web, this is the challenge we have been solving at Hunch.” Betabeat says that, “Something like 70% of items on eBay don’t have traditional metadata like product IDs.” No metadata on eBay transactions means little opportunity for eBay itself to build a sophisticated recommendation engine, and clearly the acquisition of Hunch is meant to address that.

But there’s a flip side to this equation, too, as Betabeat describes it: “eBay’s 97 million users and 200 million active listings add up to 9 petabytes of data across two billion daily page views. ‘There is only so much you can teach your system with the big academic data sets that are publicly available,’ Dixon said. ‘With eBay’s data behind us, expect Hunch to get much, much better’.”

Faster than the speed of light?

Earlier this fall, scientists at CERN said they’d clocked neutrinos traveling faster than the speed of light. Since this changes everything, there’s been an open call to other scientists to replicate the findings.

There’ve been a couple of new salvos this week: The Italian Institute for Nuclear Physics (INFN), which runs the Gran Sasso lab, has just confirmed the test results, The Economist reported. Then, another Italian site reported different findings. The ICARUS experiment said that, no, the original research hadn’t adequately accounted for the neutrino’s energy upon arrival. Recalculating the data, they argued that the neutrinos were still traveling at the speed of light — no faster.

So is the Theory of Relativity still intact? Stay tuned …

Got data news?

Feel free to email me.


tags: , , ,