The big data era is being propelled by massive amounts of unstructured data, continuously produced and stored at a decreasing cost. The increasing pace of data collection and analysis has resulted in scientific advances that are more data-driven, according to Princeton University researchers Jianqing Fan and Han Liu, and Johns Hopkins University researcher Fang Han. For example, they note expanding streams of social network data are being used to predict influenza epidemics, stock market trends, and box-office revenues for particular movies. "It is anticipated that social network data will continue to explode and be exploited for many new applications," the researchers say.
Big data researchers have focused on the development of new computational infrastructure and data-storage methods, as well as on fast algorithms that are scalable to massive data with high dimensionality. In addition, the researchers note massive sample sizes have given rise to big data that fundamentally challenges the traditional computing infrastructure. "In many applications, we need to analyze Internet-scale data containing billions or even trillions of data points, which makes even a linear pass of the whole dataset unaffordable," the researchers say. They offer Hadoop as an example of a basic software and programming infrastructure for big data processing.
View Full Article
Abstracts Copyright © 2014 Information Inc., Bethesda, Maryland, USA
No entries found