The Indian elections of 2014 saw a flurry of political startups explode. Namely, they comprised Voxta, Frrole and Simply 360. Frrole is a big data company based in Bengaluru. They have been known to filter numerous political posts on Twitter, and persuade voters.
How this essentially captures the application of big data is remarkable. To begin with, there were half a billion such Twitter posts per month. Big data begins from such a massive database.
Simply 360 went beyond Twitter posts. They presented customized analytics by spanning terabytes of data across varied media. They boasted of clientele such as Revlon, Yamaha, Target and Wipro.
Reporters remarked, saying that the usage of big data for elections were no more confined to the US elections only.
We must all remember Paul the Octopus, who supposedly predicted the outcomes of the matches of the World Cup 2010. Baidu, China’s top search engine, maintained a World Cup predictions page, and the head of Baidu’s Big Data Lab, Zhang Tong, said, that the accuracy of predictions depends “mainly on the data and the model”. While maintaining the predictions page, they had analyzed five factors: team strength, home court advantage, recent game performance, overall World Cup performance, and bookmaker odds. They aggregate the data and then, use a machine learning model to make the predictions.
In fact, another one of Baidu’s feats involved the Chinese New Year. Being the most popular holiday of China, it is also deemed to be “the largest annual mammalian migration on Earth”. At this time, 3.6 billion passenger trips are made across all modes of transportation. So, Baidu gathers data from smartphones with applications which tell locations, and register 3.5 million position requests every day. The heat map updates every four to eight hours, showing the most popular locations and travel routes.
Jersey-based Terra Seismic uses satellite data to predict major earthquakes anywhere in the world with 90% accuracy. It has often been denied that accurate earthquake predictions can be made possible. However, with big data analysis, using satellite and atmospheric data combined with statistical analysis, the claim is on the verge of being undone.
Terra Seismic’s predictions included a warning on the 22nd of February of the Indonesian island of Sumatra being subject to an earthquake of magnitude around 6.5 on the Richter scale. Indeed, on the 3rd of March, the island faced a quake of magnitude 6.4.
Oleg Elshin, CEO of Terra Seismic, mentioned, that they could predict any major earthquake (of magnitude equal to or more than 6.0) thirty days before their occurrence in any earthquake-prone country. He said, that they use satellite big data technology.
Ohio-based data science company, Farsite, accurately predicted the winners of Best Picture, Best Actor, Best Actress, Best Actor in a Supporting Role, Best Actress in a Supporting Role and Best Director for the Academy Awards of 2013. They created a statistical model, which maintained two important factors: forty years of Oscar history could help them map the trends, and momentum, popular opinion and controversy can override such trends.
Exploring the usage of big data spans across massive spectra, and will perhaps, form a database of its own. The “how to” of big data can be obtained from such examples. What they do is locate a trend, and apply real-world situations to it.
One may think, this is the ultimate machine parallel of the human act of anticipation – that we are creating something that will make our machines think.
In fact, this is what it is gradually becoming. Big data can be thought of as a machine duplicate of a brain.
Data Science Foundation in association with NASSCOM is organizing “Data Science Summit, Kolkata 2015” on 28th August at The Park, Kolkata. Something of this sort and magnitude is happening for the first time in Kolkata and we are happy that Kolkata Bloggers is a happy part of this event. We have the best of the speakers in this region coming together to raise awareness of Big Data. Do check out their website and book your calender as soon as possible.