Tuesday, December 15, 2009

Overlap in the Business Intelligence / Predictive Analytics Space

I've received considerable feedback on the post Business Intelligence vs. Business Analytics, which has also caused me to think more about the BI space and its overlap with data mining (DM) / predictive analytics (PA) / business analytics (BA). One place to look for this, of course, is with Gartner, how they define Business Intelligence, and which vendors overlap between these industries. (I think of this in much same way as I do DM; I look to data miners to define themselves and what they do rather than to other industries and how they define data mining).

I found the Gartner Magic Quadrant for Business Intelligence in 2009 here, and was very curious to understand (1) how they define BI, and which BI players are also big players in the data mining space. Answering the first question, data analysis in the BI world is defined here as comprising four parts: OLAP, visualization, scorecards, and data mining. So DM in this view is a subset of BI.

Second, the key players in the quadrant interestingly contains only a few vendors I would consider to be top data mining vendors: SAS, Oracle, IBM (Cognos), and Microsoft in the "Leaders" category, and Tibco in the visionaries category. Of these, only SAS (with Enterprise Miner) and Microsoft (SQL Server) showed up in the top 10 of the Rexer Analytics 2008 software tool survey, though Tibco showed up in the top 20 (with Tibco Spotfire Miner).

I think this emphasizes again that BI and DM/PA/BA approach analysis differently, even if the end result is the same (a scorecard, dashboard, report, or transactional decisioning system).


Will Dwinnell said...

You mention IBM/Cognos being included as a BI/data mining vendor, but I wonder for which products is this company so identified? Cognos once (a long time ago) sold data mining tools: a neural network shell and a tree induction program. I don't think Cognos had sold either for some time before being acquired by IBM. IBM used to sell a neural network program and, later, its Intelligent Miner product, but I am pretty sure both of those have been defunct for some time. That leaves the querying and OLAP tools from Cognos and I presume something similar from IBM.

Dean Abbott said...

I have poked at Cognos a couple of times during the years, but really only in the context of PA (they have some decision tree capabilities as you mentioned).

IBMs Intelligent Miner was a fine data mining tool, and once considered in the upper-tier of tools available commercially (that was in the 90s), but it was only last year that I finally worked with a company that was actually using it (but they were transitioning away from it, and the nature of my consulting with them was helping them convert processes and models built with IM into Clementine).

Have you ever used the BI aspects of Cognos? I can't comment at all on the BI capabilities myself.

Will Dwinnell said...

I worked for Cognos (pre-IBM acquisition) as a full-time employee for about 2 years. I was a post-sales consultant and set up many Cognos installations and taught many classes for them.

I found the reporting tool (GUI front end for database querying) and the OLAP too to be fairly capable. In the hands of a qualified user, these could be quite useful. My reservation, though, lies in what constitutes a "qualified user". Without some analytical and technical grounding, it is just as easy to turn out nonsense results with querying tools and OLAP tools as it is with spreadsheets. I would rate the Cognos data mining tools of that time as "weak".

Lou Bajuk-Yorgan said...

Since these ideas do overlap quite a bit, I think it makes sense to picture them as a spectrum, rather than one or the other being the umbrella under which everything else must fit. One the one end, you have business intelligence, dashboards and reporting, which are typically more static and backward-looking. These reports get deployed to large numbers of users—who may or may not actually read them.

On the other end, you have statistics and data mining—techniques which provide a tremendous ability for making sense of your data, predicting future behavior, and helping you focus on the relevant trees in the forest of data. These techniques are typically limited to a smaller cadre of the most analytically-sophisticated users.

In the middle, you have business “analytics”, where a large number of people in both business and technical roles need to analyze data try to extract information to make decisions. The data may downloaded from a BI system (when the BI report doesn’t provide the user with the insight they need), or it may be analytically-enhanced with the results of data mining (enabling the user to focus on the most relevant data, perhaps with value-added information like customer scores or segmentation information). These users aren’t typically statisticians or data miners, so they struggle with spreadsheets, or (better yet) use visual analytics to get the insight they need (such as provided by an application like TIBCO Spotfire, for example).

Ralph Winters said...

Note that the Gartner definitions are based upon their heading Marketing/Descriptions. Query and Reporting and Ad hoc querying, which encompasses maybe 85-95% of all BI analytics in corporate organizations today is grouped in the Information Delivery section rather than in the Analytics sections.

-Ralph Winters

Website said...

hey nice information for me,thanks for sharing this information.I have to sure to bookmark this blog.

product data mining