A Glance into the Future with Big Data and Super Computers

We are always interested in knowing about the blurred mysteries of the future. In fantasies or fictions and in some historical contexts some where we have witnessed humans acquiring skills to make future predictions and write prophecies. To know about future, interest everyone alike. But the question is whether it is possible today to predict future with all the latest technologies available at our disposal?

The answer to this question is YES, with right approach it is actually possible to predict certain events of the future accurately and to do so BIG DATA has a major role to play. This article throws light on BIG DATA and present some successful results on how by studying some specific patterns shown by the Big Data with the help of super computers it is actually possible to predict the future events.

An introduction to Big Data
Big Data, as name suggests, is a collection of unimaginably huge amount of data in various forms like texts, pictures or videos. With the exponential growth of Internet, we are generating gigantic amounts of data each day which contributes the ocean of Big Data. There are innumerable numbers of activities going on Internet each second. With every Facebook update, each Tweet posted, every picture or video uploaded and with internet browsing we are generating huge amount of data that collectively make Big Data. An estimate by IBM suggests that 2.5 quintillion bytes (i.e. ten raised to power 18 bytes) of data is being generated everyday by us and that is how Big Data is growing. In fact, in last two years about 90% of world's data has been generated.

How can Big Data make a good prophesy?
As we understand that BIG DATA is a huge amount of data that contains a lots and lots of information. If we have some way to process this enormous data systematically it is possible to predict future events. It is like weather forecasting! Today with super computers that have enormous computing power, it is possible to process Big Data to make future predictions like political upsurge, war, natural calamities etc. and there are results to show for it. Attempts in this direction were started in 2011, when an 8.2 teraflops supercomputer, called "Nautilus" (placed at University of Tennessee) successfully predicted the conditions for a big revolution in countries like Tunisia and Libya, and the fall of Hosni Mubarak from power in Egypt.

The IBM Nautilus Supercomputer

How was it done technically?
The methodology was to feed a huge database of past year's news to the machine and read the patterns generated henceforth. Data scientist Kalev Leetaru is one of the foremost people in the field of "predictive supercomputing". The data collection (fed to super computer) prepared by Leetaru and his team consisted of over one hundred million articles from the media around the world over a time span of 30 years. Each news entity was translated and tagged for geography and tone. They analyzed the processed data on the supercomputer Nautilus creating a network with ten billion items connected by one hundred trillion semantic relationships. The outcomes of this magnificent study (named Culturomics 2.0) were impressive – giving signals for a great degree of predictive ability of Nautilus. The events that could be successfully predicted included the revolutions in Tunisia, Egypt and Libya, including the fall of the then Egyptian President Hosni Mubarak from power.

It is seen that the analysis or the results of the experiment are based on the large amount to data being fed to the computers or in other words the results are the conclusion of information known before hand. It will be interesting to see how this technology can be implemented in real time applications where the events of today are tracked to predict events of tomorrow.

-By Anirudh Khanna - Student 3rd Year CSE, Chitkara University Punjab