Intel Software Adrenaline

Perform Predictive Analytics and Interactive Queries on Big Data

Issue link:

Contents of this Issue


Page 0 of 7

White Paper Intel® Xeon® Processor- and Intel Atom™ Processor-Based Servers Big Data Analytics Perform Predictive Analytics and Interactive Queries on Big Data Abstract Recent innovations in data warehousing and business analytics dramatically increase the capability and potential value of today's massive, diverse, and often fast-moving data flows. Companies now perform interactive queries and predictive analytics using all available data, including operational data and the huge amounts of poly-structured data available from logs, social networks, sensors, and many other sources. In this white paper, we define a practical, cost-effective infrastructure for supporting data-driven decision-making on an enterprise scale. These solutions build on the hybrid IT infrastructure introduced in an earlier Intel paper, "Extract, Transform, and Load Big Data with Apache Hadoop.*" Overview Parviz Peiravi, Principal Architect, Intel Enterprise Solution Sales Ajay Chandramouly, Big Data Domain Owner, Intel IT Chandhu Yalla, Big Data Engineering Manager, Intel IT Moty Fania, Principal Engineer, Big Data/Advanced Analytics, Intel IT The torrents of data flowing into businesses today can fuel new insights and actions in near-real-time. Organizations ready for that opportunity stand to gain a substantial competitive advantage. According to a recent article in Harvard Business Review by Andrew McAfee and Eric Brynjolfsson, "The more companies characterized themselves as data-driven, the better they performed on objective measures of financial and operational results. In particular, companies in the top third of their industry in the use of data-driven decision making were, on average, 5 percent more productive and 6 percent more profitable than their competitors."1 Most IT decision-makers are well aware of the value big data analytics offers to their business. In an Intel survey of 200 IT professionals, 46 percent considered improving analytics capabilities to be one of their top priorities, and 90 percent rated the importance of improving analytics either 5 or 4, on a scale of 1 to 5, where 5 is "most important."2 The traditional approach to analysis is to copy data from operational systems into a data warehouse for queries and analysis. While that model still applies to many scenarios, companies now integrate more data—and more diverse data—into their analytics environments, including documents, multimedia files, operational logs, social networking posts, Internet click-streams, sensor measurements, and more. Businesses combine historical data with fresh operational data and high-volume streaming data to achieve a 360 degree view of their customers and their internal operations. Many companies also are shortening the time it takes to analyze large data sets, so they can integrate analytics into time-sensitive business processes. All of this can be accomplished today using proven methods, but the overwhelming volumes, continuous generation, and poly-structured nature of big data requires upgrades and additions to traditional infrastructure.

Articles in this issue

Links on this page

view archives of Intel Software Adrenaline - Perform Predictive Analytics and Interactive Queries on Big Data