I Didn’t Know Big Data Could Do That!
You’ve probably heard about some of the common uses for big data technologies–how phone companies reduce customer “churn” with Hadoop, how spy agencies find terrorists with pattern recognition, and how biologists speed drug discovery with graph databases. But the power of data analytic technologies goes far beyond these established use cases, and can be used in genealogy, law, and even pop music.
Organizations today are finding creative and different ways to harness the power of big data analytics beyond the traditional ones. Here are five non-intuitive uses of big data technologies in the real world that will make you say “I didn’t know big data could do that!”
Predict Outcomes of Lawsuits
If you’ve ever been party to a lawsuit, you understand how tough it can be to get the right lawyer on your side. You may explain your case to five or six attorneys before finding a good fit. Now, with big data, those legal eagles are flipping the tables and analyzing lawsuits en masse in pursuit of the cases with the best prospects for big payouts.
A new product from LexisNexis called MedMal Navigator can significantly shorten the time required to determine whether or not to take a medical malpractice case. The software, which was unveiled a year ago, compares years’ worth of case law with the specific facts in a case, allowing lawyers “to determine in 20 minutes–versus 20 days–if a case is worth taking on,” according to a press release.
Another company, called Lex Machina, does the same thing, but for intellectual property cases. The company combs through all available sources of data on the Web, catalogs it, and then uses that database to determine how new cases will fare, based on several variables. “We mine litigation data, revealing insights never before available about judges, lawyers, parties, and patents, culled from millions of pages of IP litigation information,” the company says on its website.
Make Craigslist Safer
Buying things on Craigslist has become commonplace in many parts of the country, but that doesn’t mean that it’s safe. According to a 2011 study by the AIM Group, Craigslist was linked with 330 crimes, including 12 murders and 105 robberies or assaults, over a period of 12 months across the country. The study said Craigslist was “a cesspool of crime.”
A company called ByOwner.com is hoping to harness the power of social media to make the popular classified ad website safer. Last year the company began testing a service that taps into Facebook’s huge graph database of people to provide a way for prospective Craigslist shoppers to authenticate the identity of the people they are thinking of buying stuff from on Craigslist.
The ByOwner.com service leverages Facebook’s APIs to determine whether the seller is a real person, which is done by analyzing the seller’s connections. “The more you know about the other person, the more safety and security you have,” ByOwner.com founder Ramish Radhakrishnan tells Datanami.
Spot Pregnant Women and Criminals
What you buy says a lot about who you are and what you do. But did you know that your purchases can also tell a retailer if you’re pregnant or a criminal?
Target has access to the personally identifiable information (PII) of hundreds of millions of people. (And the retailer also has trouble keeping that information secure, but that’s another story.) The company uses big data analytics to pull useful information out of its database, so that it can more effectively market goods to customers and generate bigger profits.
As Forbes reported a couple of years ago, the company mines its data store looking for certain purchases that are strong indicators that a woman is pregnant. For example, did you know that pregnant women are big buyers of scented lotion during their second trimester? Or that cotton balls and washcloths are frequent purchasers late in the third quarter? It’s true, and Target’s sharp algorithm-writers use these carefully curated facts to help it reel in mothers-to-be and make them loyal Target shoppers before the real spending begins, i.e. the baby is born.
What you buy can also tip retailers off to a less wholesome clientele. TIBCO had a customer in Europe several years back that was trying to figure out what purchases indicated a customer was using a stolen or fraudulent credit card. “We looked at it as a math problem,” TIBCO CEO Vivek Ranadivé said during last fall’s TUCON conference. “We found that if you bought razors, champagne, and diapers, it was probably a stolen credit card.”
It was pretty simple to explain these purchases, Ranadivé said. “Razors blades are big ticket items that are easy to pawn off. That makes sense. Ditto for champagne,” he said. “But why diapers? Maybe he was trying to look like a good guy, or a dad. By converting it into a math problem, we were able to find something without having to know the why.”
Discover Your Ancestors from Spit
The folks at Ancestry.com have built their own big IT infrastructure to help its 2.7 million subscribers discover their ancestors, explore their genealogy, and share what they find. According to an in-depth report on Ancestry.com’s hyper-scale infrastructure on Datanami’s sister publication EnterpriseTech, one of the drivers of growth at the company is the AncestryDNA services that it launched nearly two years ago.
The AncestryDNA service is fairly simple from the customers’ perspective. First, spit into a vial, then mail it back to Ancestry.com, which then analyzes it and reports the results back to the user. For about $100, anybody can have their DNA analyzed. But what makes Ancestry.com’s service special is the Hadoop cluster it uses to match snippets of customers’ DNA to others. This service helps customers find cousins and track their genetic lineage.
Identify the Next Hit Song
Big music companies have been hammering away at the wall between art and science for decades. They’ve used statistical analyses of album sales and radio plays to help determine which acts they should invest in. Now the Internet is changing the dynamics thanks to the big data generated by social media, online music stores, and streaming music sites.
Companies like Next Big Sound are gobbling up every piece of available data that can signal whether a particular act, song, or album is trending up or trending down. The outfit has confirmed what may seem obvious–that increased social media activity, such as Facebook likes and Twitter followers, does indeed correlate with sales on places like iTunes. This information helps artists and music companies predict how songs or albums are likely to succeed in the market.