IBM just announced a number of upgrades to its predictive analytics technologies. They focus on squeezing business insight from sources of information that many enterprises are still only trying to manage. 

The releases were announced yesterday at IBM’s Information on Demand Forum in Las Vegas. Generally speaking, they consist of a number of upgrades that will apply cognitive computing to analytics technologies, giving companies an easier way of forecasting trends.

Analytics in Context

So what’s new here? Leaving aside the fact that there are a number of upgrades, what IBM is saying with this release is that it is time to move beyond managing big data and time to start using it for business value.

The key difference with other big data applications, IBM  claims, is that its technology goes beyond the mere management of big data and breaks it down into useful information. It is also announcing the expansion of IBM’s BLU Acceleration portfolio of products.

Big enterprises have problems, IBM contends. With the growth in the number of computing environments, including the cloud and mobile, enterprises are creating around 1.3 terabytes (TB) of data daily. How much data is that? Let's digress for a moment.

Joel Lee, in a post on Memory Sizes Explained, notes that a gigabyte (GB) is 1,024 MB. That’s equivalent to 10 yards of books standing side by side or 200 MP3 songs running at 5 minutes each. A TB equals 1,024 GBs. He explains:

You would need approximately 1,500 CD-ROMs to match the capacity of a 1 TB hard drive. Now, let’s say that the dimensions of a CD case are 142 mm x 125 mm x 8 mm (yes, I measured). If you stacked 1,500 CD cases on top of one another — and kept it from toppling over — it would reach a height of 12 meters or 39 feet.

Back to our point. Enterprises create massive amounts of data — including log files, software error alters and network configuration updates, resulting in more than one million systems alerts every day.

Some of these alerts are important. Some are  irrelevant, particularly for customer facing workers or even systems administrators, who can end up sinking under a sea of pointless alerts.

However, taking some of the lessons learnt through the Watson computing project, IBM claims its technologies can differentiate the context of the user questions and queries and cut through irrelevant information to provide accurate responses.

Predictive Analytics Upgrades

The first upgrade IBM announced involves its SmartCloud Analytics — Predictive Insights, technology that can "provide early problem detection to predict application or middleware problems before they impact service."

For business users, the value is that the upgrades enable the software to learn and "sense" how an IT network is operating and then upgrade or change it as necessary. The software identifies changes to the way users are accessing and using the network, no matter how subtle those changes are, IBM reported.

In theory, as external business conditions change, the software can identify poor system configurations that do not respond to the way users are working at a given time.

IBM is also applying these analytics to the way enterprises are storing information with a new version of the SmartCloud Storage Virtual Center. This will effectively automate the storage of information in appropriate locations, as well as facilitate the move to cloud storage.

IBM does this by analyzing data usage patterns and identifying the best way of storing information for users. Then it can move the data, without disrupting access to business applications that depend on the data.

Analytics Improvements

There are a number of upgrades that are worth noting here, including:

  • InfoSphere Data Explorer: A new big data exploration tool for enabling exploration of all cloud-based and enterprise systems, and the discovery of information no matter where it resides in the enterprise
  • InfoSphere Data Privacy for Hadoop: allows clients to anonymize data in Hadoop, NoSQL and relational systems.
  • IBM PureData System for Hadoop: This simplifies the deployment of Hadoop systems and enables users to get up and running in hours. Other enhancements include built-in archiving tools, simplified administration and better security.

IBM has also announced it is expanding its BLU Acceleration portfolio "to help clients harness big data and analytics," including an early access preview of BLU Acceleration for Cloud. This means enterprises will be able to use IBM’s by BLU Acceleration in-memory database and business analytics technologies in the cloud.