PARIS - Computer services giant Atos called 2012 the year of evangelization of "Big Data." Now, in 2013, it's time we may all be converted.
Big Data is the massive volume of digital data that is so large to process that it needs additional management tools and database processing applications. It is much more than a concept; it’s a whole new technology that promises to “revolutionize our everyday lives.”
The term “Big Data” originated from a simple observation. Every day around the world, 2.5 exabytes of data are generated. This data comes from everywhere: social networks, photos and videos posted on the Internet, GPS coordinates from smart phones, weather statistics from every corner of the earth, banking transactions, and so on.
This massive amount of information only reveals its full potential after being processed, analyzed and cross-referenced. “Stocking data is quite meaningless, what makes a difference is the way we use the data to turn it into something useful,” explains Yann Cohen, head of Greenplum Europe, a Big Data provider.
“Internet changed the world in the 1990s, Big Data will change the world in the 2010s,” anticipates EMC technical director of world leader in storage and Big Data guru, Chuck Hollis. “Big Data offers us a new way of thinking of seeing things. It’s the ‘new oil’ of the information age.”
Predicting heart attacks
In practical terms, how does this Big Data thing work? There are many examples of major innovations that come from cross-referencing data collections. For instance, if you input the eating habits or health history of someone into a computer, the computer could – in a matter of seconds – predict when he or she will have a heart attack. Using the same concept, retail shops could use their peoples’ buying history to provide custom promotions to potential clients walking past their shops.
The implications of Big Data are numerous – in theory at least. For the moment, the analysis of massive amounts of data mostly concerns companies who use it as a way to analyze their clients’ consumer habits. “A company like Apple can follow in real-time how many iPads it is selling, and compare that to what is being said on social networks,” explains Yann Cohen.
“Being able to analyze what people are thinking is capital. Using semantic tools, we are able to sort the positive reviews from the negative ones on the social networks, even the sarcastic ones!” confirms Serge Boulet, marketing director from SAS, world leader of Big Data for companies. “The objective is to identify potentially negative signals as quickly as we can, so we can respond.”
Last October, Canadian company RIM (Research in Motion) experienced a major outage on its BlackBerry network. It took the brand a few days to issue a public statement, and in the meantime, its shares lost 60% of their value, not to mention its image. “Situations like these demand immediate responses, and this is where Big Data real-time monitoring comes in,” explains Yann Cohen.
Another application of Big Data is price fluctuations. One of SAS’s American clients, a textile retailer is using Big Data to adapt its prices. “In the textile industry, the constant challenge is setting the right price according to size, color, localization, weather, etc.,” says Serge Boulet. “Big Data enabled us to set up a system where thanks to electronic tags, every night we can change all our prices within minutes. Thanks to this process, the shop can offer markdowns at the perfect time and adapt its prices to optimize sales.”
The problem is the public’s acceptance. Big Data cannot be allowed to turn into “Big Brother.” “The users’ trust is crucial,” says Chuck Hollis. “To get the best medical treatment, we give our doctor as much information as we can, because we trust him. Big Data needs to convince its users it can be trusted in the same way.” He adds: “Every new technology creates new fears. Fire, electricity, the Internet, Big Data. With good there is always bad, we just need to keep that in check.”
In France, unlike the U.S., the CNIL (the French data privacy authority) says personal data must remain anonymous. When Internet users’ information is cross-referenced, it is done not with single users but groups of users.
Cross-referencing raises major issues in regard to personal data protection. The CNIL recently asked Google to better inform its users about the way it collects their data, stores it and cross-references it.
Sometimes, Big Data gets it right, as the last American presidential election proved. Statistician Nate Silver was able to predict the exact results of 49 out of 50 states, making it to Time’s 100 most influential people of the year at the age of 34. Are stats better than polls? That is certainly what came out of the last U.S. presidential election – which ended up being called the “the Big Data election.”
Campaign staffs largely used the process to target potential voters. “Before this, people would make somewhat random calls – but this year, I received no phone call, no email from any of the candidates,” notes Chuck Hollis. “Well, I’m a 53-year-old Republican, but my 22-year-old daughter who was voting for the first time got calls from all the parties through the whole campaign!”
“We are usually six to nine months behind the U.S.,” says EMC France CEO Jean-Michel Giordanengo. “But there is no doubt that Big Data will rule in 2013, as a competitive tool for businesses who are looking to provide their customers with added services.” Digital data will certainly keep on increasing, indefinitely. After the social networks, more connected items will add their data to the pile: cars, electricity meters, TVs, etc.
“It’s not a trend, it’s a revolution in the making,” concludes Serge Boulet.