SPSS updates data mining tool

Government agencies seeking to mine data for purposes ranging from detecting

fraud to improving customer service have a new tool at their disposal.

SPSS Inc. began shipping Clementine 6.0, an update of its flagship data

mining solution, this month. Clementine is designed to help the public sector

and businesses gain an in-depth understanding of data to help them attract

and retain customers, detect fraud, reduce risk and improve government service

delivery, said Peter Caron, senior marketing manager for the Chicago-based

company.

"Clementine is a data-mining workbench that helps users make predictions

based on data," Caron said. "For the government, it helps predict what citizens

are going to do."

Federal, state and local agencies are using Clementine in many ways:

* The Defense Finance and Accounting Service is testing data mining

as a way to discover billing errors and fraud. DFAS' vendor pay branch uses

data mining to search through 2.5 million financial transactions for inaccurate

charges. The software enables DFAS to screen each transaction for 80 elements,

including the item bought, the price and how it compares with previous purchases.

* The Internal Revenue Service uses Clementine to schedule its workers

in call centers and elsewhere to help improve customer service.

* The Georgia Department of Education recently began using Clementine

to help determine the best curriculum structure for its schools by analyzing

whether students tend to perform better if their classes flow in a certain

sequence.

"The key to data mining is delivering personalization, predicting what's

going to happen in the future," Caron said. "That predictive aspect leads

to smarter business decisions."

Fiona McKenna, industry marketing manager for the public sector at SPSS,

added: "It helps determine why something is occurring and allows an agency

to prepare itself for X, Y and Z to occur," she said. "That provides a real

value for citizens who find government confusing."