Data Mining Made Simple

SPSS updates add graphical capabilities for ease of use.

SPSS Inc. is trying to shed the complexity inherent in data mining software with upgrades of its SPSS for Windows and SPSS Server for Windows NT/2000/2003, chiefly targeting usability and presentation improvements.

The Chicago-based companys data mining software is typically used by statisticians for complex analysis, with SPSS Clementine product targeted for less technical users. But in Version 12.0 of the high-end offerings, due this week, SPSS is trying to broaden the appeal of its base software, primarily through the use of GUI enhancements.

SPSS 12.0 features a new presentation graphics system, used for adding graphs to reports generated by the software. Enhancements include improved production graphics, expanded control over chart layout at the time of creation and customizable chart templates, SPSS officials said.

"The products come a long way since [Version] 11.5," said William Bailey, principal statistician at WMB & Associates, in Apopka, Fla. "Theyve taken a lot of flak for their graphical modeling capabilities in the past."

Bailey said the new graphical capabilities should make the product more "palatable" to beginning users. "Theyve done a lot of things to make the product more user-friendly to a larger audience," he said. "Im a core stats user, but now the doors open to an audience less stringent than I am, more of a general statistics user. But behind the scenes, theres still hard-core data analysis going on."

In addition to a better presentation layer, SPSS 12.0 includes capabilities to speed the data analysis process and give users more control over that process. These capabilities include support for longer variable names to describe data files. Previous versions of the software supported only eight-character names, Bailey said.

Users can also assign parameters for identifying and filtering duplicate records. In addition, they will get better capabilities for sorting and breaking down data into component groups, such as breaking ages into demographic groups.

Data sets can be made read-only to protect their integrity, and users can add notes to data files for tracking purposes.

SPSS 12.0 also includes a new add-on module called Complex Samples, which allows users such as researchers and social scientists to incorporate the sample survey methodology they use into their SPSS data analysis.

SPSS 12.0 is available both as a stand-alone desktop product and in client/server format, using SPSS Server.