Current trends in predictive analytics of big data. (1st January 2014)

• Record Type: Journal Article
• Authors: Wlodarczyk, Tomasz Wiktor; Hacker, Thomas J.
• Is Part Of: International journal of big data intelligence. Volume 1:Number 3(2014)
• Page Start: 172
• Page End: 180
• Publication Date: 2014-01-01
• Subjects: big data -- data intensive -- predictive analytics -- predictive modelling -- elemental data -- time series; Big data -- Periodicals; 005.705
• Languages: English
• ISSNs: 2053-1389

Abstract

Predictive analytics is a driving force motivating considerable interest in big data. Although there is clear interest in big data, the adoption rate of analytical techniques fuelled by big data that can extract knowledge and value from these data is less well understood. In this paper, we present a quantitative analysis of trends in publications related to predictive analytics, predictive modelling, big data and data intensive computing. Our evaluation shows an increasing popularity of big data in scientific publications, with ten-fold increase in the last three years. Concomitantly, we find that predictive analytics are connected with this trend, with two-fold increase in the last three years, but also a seven-fold increase in the same period when used in context with big data. We also classify the main application domains for big data and predictive analytics. Contrary to popular belief that big data is focused primarily on social media and business intelligence, our analysis found that almost half of scientific publications using predictive analytics were in healthcare, smart services, the internet of things, and weather and environment. Our results indicate the early adoption of big data-based analytics in these domains.