IBM_logo_2009.jpg IBM (news, site) has just released an upgrade to the SPSS predictive analytics tool that will enable users to analyze text gathered from social media sources, including Twitter, Facebook, blogs, wikis and RSS feeds.

The upgrade represents a considerable advance in the SPSS Modeler data mining and text analytics solution with abilities that stretch as far as detecting the context of the words used. Data from the social web can also be merged with internal data to create even more accurate intelligence about consumers, IBM says.

With its new ability to understand emoticons, slang and other social media terminology, as well as jargon from specific industries, it includes taxonomies from 180 verticals ranging from Life Sciences to Financial, Insurance and Consumer Electronics. IBM SPSS Modeler 14 Premium Edition is designed to analyze any information that gives companies insights into how their products are being received.


This is the first major upgrade to Modeler since IBM bought out predictive analytics company SPSS for US$ 1.2 billion last October.

At the time IBM said the acquisition would further expand its Information on Demand (IOD) software portfolio and business analytics capabilities. Forrester is interpreting the move as the first salvo in a battle for supremacy in the business intelligence market.

SPSS Modeler 14

And with this upgrade, IBM has added to its social and business analytics capabilities and really upped the ante for competitors. Modeler 14 Premium Edition, by analyzing sentiment in social media, can predict what consumers are interested in, where they are going to go, and what practices will keep them interested.

IBM SPSS Modeler 14 Premium Edition workbench

It comes with new semantic networks and more than 400,000 terms across most verticals as well as 100,000 synonyms and thousands of brands enabling users to link sentiment and products without having to build up their own definitions within those 180 taxonomies.

SPSS_Add text.jpg

In this respect, IBM cites the banking industry as an example where the semantic network covering it can distinguish between much of the terminology governing mortgage lending. It knows, for example, that a “floating rate” is the same as a mortgage loan and that variable and adjustable mortgage rates are effectively the same thing.

Business Intelligent Decisions

The bottom line here is better business decision making. Using Natural Language Processing it takes key ideas, opinions and expressed sentiment from the web, including survey data and text and integrates that data into models designed to predict customer reaction as well as recommending better business decisions.

By putting this together with the structured data that already exists in the enterprise, users can uncover relationships between ideas and consumer actions so as to build better understanding of customer groups and how they are reacting in a particular channel.

IBM BI Expansion

And as to why IBM went down this route in the first place, at the time of the SPSS acquisition Forrester (news, site) cited three reasons:

  1. Continued growth of its Information On Demand drive -- its smallest business area -- that has seen it buy out Ascential Software, FileNet and Cognos. IBM’s vision of a Smarter Planet combines instrumentation, interconnectivity and intelligence. Without SPSS, Forrester says, IBM would not be able to deliver on this.
  2. Strategic move to counteract the acquisition of Sun Microsystems by Oracle and put itself ahead of the other big guns at a services level.
  3. Improves market placement by putting itself in a better position than any other vendor to deliver integrated smart solutions in the future, from physical world awareness to decision support tools.

The latest version of IBM SPSS Modeler data mining and text analytics workbench is now available worldwide. Text analytics workbench is only available in IBM SPSS Modeler Premium edition.