Hedge fund traders bloomberg terminal

236 Harness the Power of Big Data: The IBM Big Data Platform

A lot of clients are on the right track when they ask us about integration points from their existing technology investments and their new Big Data projects; indeed, that’s always encouraging to hear, because it means that the client is thinking beyond the “science experiment” phase that rarely produces value. Note that there are many more integration points than we can possibly cover in this chapter, so we encourage you to use this chapter as a jumping-off point, rather than a definitive discussion of integration options and patterns.

IBM Cognos Real Time Monitoring (Cognos RTM), a component of IBM Cognos Enterprise, is software that provides visualization and analysis on real-time streaming analytics from Streams. Visualization is one of the major challenges that Big Data brings to business analysts; in fact, some universities today actually offer degrees in Big Data visualization. Throughout this book we’ve emphasized how data in motion is a differentiator for the IBM Big Data platform, because velocity (one of the four Big Data characteristics) isn’t being

Integrating Big Data in the Enterprise 237

SPSS

IBM SPSS software is a broadly deployed and sophisticated predictive ana-lytics portfolio. It’s no surprise that customers want to leverage SPSS to interact with the IBM Big Data platform. At-rest modeling is reasonably well understood, but what if you want to take predictive analytics to the extreme—to make predictions as the data comes into being, be it at a point of sale, a hedge fund trader’s Bloomberg terminal, or even emergency room monitoring equipment? After all, the whole point of predictive modeling is to improve outcomes. The combination of SPSS and Streams is the answer, as it gives you the ability to build predictive models at rest which are then scored using the data in motion. To support this powerful usage pattern, SPSS predictive models can be directly exported to the Streams run-time environment. Additionally, Streams supports models defined by the