As I write this column, the new NFL season is about to begin. At this time of year, my attention turns to my beloved San Francisco 49ers, led by Steve Young, sometimes quarterback, sometimes lawyer and full-time computer geek. It occurs to me that Steve and the Niners would ensure a Super Bowl win this season if they would get rid of those picks and shovels and trade them in on some brand new data mining tools. So I e-mailed ol' Steve (he's 38 ancient by both football and geek standards) and suggested that he take a look at http://www.SPSS.com/.
Any of you readers out there who may have suffered through a course in probability theory or market research during the last three decades or so are no doubt familiar with SPSS, one of the leading purveyors of statistical analysis software. The SPSS folks tell me that they have sold their stuff to something like a quarter-million organizations for stat analysis stuff. Not content to sit on those cash-cow licenses, SPSS is investing beaucoup resources into data mining.
Most of us are aware that quants are the only people capable of getting anything useful out of most of the current data mining technology. If data mining were easy, everyone would be doing it right? Compounding the problem, might-be users are confronted by a bewildering array of choices, each one touted as the magic bullet that will enable you to blow away your competition in milliseconds.
To its credit, SPSS is striving mightily to take the high road to data mining. Its dual themes are scalability and deployment. "What the heck does that mean?" you may well ask. I'll play front man for SPSS for a minute and try to sort it out for you.
We all know that the ability of a warehouse to scale up to accommodate an ever-increasing amount of data is critical to the success of business intelligence (BI) solutions. But merely allowing the warehouse to scale up isn't worth much if one can't analyze all the data together. Many of today's front-end analytical tools not just data mining tools can't handle really big data sets, thereby limiting the utility of the warehouse. By addressing the issue of scalability as a core thread in its data mining tools, SPSS is, in effect, saying that it stands ready to efficiently handle as much data as you can throw at it.
Deployment has to do with getting the results of data analysis (inclusive of data mining) to the end user or decision-maker. SPSS has no fewer than six products it classes as deployment tools, and more are promised. Some of these tools focus on collaboration, enabling, for example, managers and staff to work on the data analysis process together. SPSS deployment solutions also permit a solution, designed by an analyst, to be replicated and shared across an organization so that people who have only very basic analysis skills can use it.
So that's the good news. The bad news is that SPSS offers such a bewildering array of tools that it is difficult to understand which ones deserve your attention and which ones don't. No fewer than thirty-three products are listed on the SPSS Web site. Many of them are complementary. Many are not. For my money, the company needs to do a much better job of presenting its wares. I've been privy to two kimono-opening, analyst-only presentations designed to provide the lowdown on the company's plans and programs, and I am still a bit confused. Maybe because I'm not the quant I used to be.
To help you sort through SPSS' offerings, the company groups its products into four areas: Reporting, Classification, Forecasting and Visual Data Mining. Each area has specific products associated with it, and all interrelate through or take advantage of SPSS' general-purpose analysis and data preparation tools.
One thing that will help SPSS in its quest for data mining dominance is that it is working closely with major services and software vendors such as IBM, so that its technologies will be embedded in end-to-end solutions, effectively isolating the user from the nuts and bolts. Presumably, such partners will eliminate a lot of the pain and suffering associated with having to decide how to make the punishment fit the crime.
I'm told that the company intends to streamline its product line and simplify the manner with which its products are presented and sold to you and me. This will certainly be a step in the right direction. Nonetheless, the Millennium Super Bowl is just around the corner, and I'm advising ol' Steve to get with it ASAP.
Register or login for access to this item and much more
All Information Management content is archived after seven days.
Community members receive:
- All recent and archived articles
- Conference offers and updates
- A full menu of enewsletter options
- Web seminars, white papers, ebooks
Already have an account? Log In
Don't have an account? Register for Free Unlimited Access