April 2014

Compliments of Platfora Inc.

Download pdf

Big data analytics is an essential tool in understanding and driving today’s combined online and physical business world. However, there exists considerable confusion about how traditional business data (e.g. POS, inventory, shipping, etc.) can be combined with modern digital data sources. Add the reality of a still evolving technological environment, and the outcome is confusion for many aspiring analysts. The aim of this paper is to demystify the situation. 

Abstract

Big data analytics is an essential tool in understanding and driving today’s combined online and physical business world. However, there exists considerable confusion about how traditional business data (e.g. POS, inventory, shipping, etc.) can be combined with modern digital data sources. Add the reality of a still evolving technological environment, and the outcome is confusion for many aspiring analysts. The aim of this paper is to demystify the situation.

We begin with a brief summary of the technical roadblocks to adoption of big data analytics directly on today’s most popular big data storage and processing platform, Hadoop. In contrast, three successful use cases, based on real implementations across different industries, demonstrate what is possible and where real benefits lie. Their success stems from advances in agile context creation and timely, graphical analysis by business users, as well as consolidating data from both traditional and so-called “œunstructured” data sources in Hadoop in their raw, original form.

This white paper also offers a useful model of the modern data landscape, breaking it up into three distinct areas: process-mediated data, machine-generated data and human-sourced information. Understanding these three data types is a foundation for advancing big data analytics and choosing an architectural approach to its implementation.