Big Data: All Under The Hood ?

In 1995, the EU adopted a law that imposes restrictions on the collection of any personal information. In this case, personal information declared any information that could identify, directly or indirectly, the identity of the person. Probably, lawmakers were thinking of the information such as, say, a personal identification number, and other important bureaucratic documents that need to be protected from the uncontrolled proliferation network. But now the information that falls under this definition much more. 18 years ago it was impossible to imagine the amount of information that are now produced by the network every day. One major type of portal youtube now generates more traffic than the entire Internet 10 years ago. Rules written recently become feasible for reasons both technical and legal nature.

What happened? First, the volumes of data that are created each year grow and grow exponentially. In 2012 this figure was 2.8 zettabyte, and according to analysts, by 2015 it will double. Nearly three-quarters of this volume is created by individuals, when they create, copy, transfer files. Internet user creates 1.8 terabytes of data per year, or 5 GB daily. This is including downloadable video files, text files, e-mail, and proprietary information, which inevitably creates any computer connected to the Internet. But there are still smart phones, tablets, TVs, cars ... In the future, the number of connected to a worldwide network of devices will only increase. 

Much of this data is not visible to users and seems to depersonalize. In fact, what can be said about the information and how much time you dwell on that site? Separately - a little bit, but if it is connected and processed along with hundreds of other parameters, such as, for example - one view movies, data about your location, disclosed your phone, speech patterns, especially these and many other features you can select of the thousands of other people. It's not the distinctiveness of such signs, and in the total picture, which they disclose. Previously, such a comparison of images of the hundreds of mosaic fragments scattered over the network was not possible, costs are too high for computing power to process massive amounts of data in ESDS data centers. 

Anyway, it was so early, before the discovery of methods for processing Big Data. Practice shows that the more and more diverse array of data, the harder it is to depersonalize. Modern science data can be used to identify any information virtually. The identification of the person is not an end in itself, the main purpose of commercial big data - this prediction among clients, existing and potential. With this noble aim and accumulate massive amounts of information not only accumulate, but are bought and sold. With the development of Big Data, this area is becoming increasingly commercialized and specialized. There are some companies that collect volumes of information, collated and reselling. 

However, this collection of information may seem old-fashioned method compared to how it works, let's say, Facebook. It did not have anything to collect, users themselves will bring everything necessary platter. According to the documentation on the IPO, Facebook holds about 111 megabytes of photos and video for each member of which there are more than a billion. And that's not all. Remaining text messages, "husky", addresses of computers from which you access, meta-information (tags), and more. And separately amounts of data collected by Facebook would provide a lot of value. In February, Facebook announced a deal with Axiom, one of the conditions of which will be exchanged and merging the user databases, allowing you to connect an array of data about the behavior of people in the reality of their behavior online, which will make more accurate predictions. 

Such a database would include, according to some estimates, up to 90% of social profiles created by the Americans. In other countries, this figure will be lower but still significant. Believed that such arrays are "anonymized" before treatment, but the more personal information they contain, the more likely that either such action is taken, in reality, or they are meaningless. For example, even if the mobile operators companies anonymizing data on the movement of its customers before the sale (ie, take away the phone numbers), it is still using the algorithms of Big Data, which take into account a lot of indirect measures, 4 points location is just enough, to link these anonymous data with a real person. 

further increase in the volume of data, it will be possible to predict the future behavior of people with ever increasing accuracy. Of course, the single unique events to predict such programs are unlikely to be able to, but the daily behavior is, in fact, it usually follows the same patterns and easily predictable. It is easy to imagine all sorts of commercial applications of such information. Something like Google is doing with his project Now, although it is not known exactly which data is collected and how exactly Now he makes predictions. But is it possible to escape from the electronic surveillance? Of course it is. Much of the data is obtained by using the Internet, cell phones, and other benefits of "smart" civilization. 

Goods that have learned to think of themselves. Who knows what they have in mind, and in the best interests of the owner whether what they are doing? But we all participate in the global race for the increase in labor productivity, which first receive it. In the end, we all have to get used to this, and will not even represent how we lived before. But the man from the past could be greatly surprised to see how many of us are willing to tell myself. Or even frightened Or cautious? I do not know. Perhaps a combination of both. The changes seem to be striking when looking at them on a scale years. But living every day, you do not notice them.