SPSS Ships AnswerTree for Discovering Segments, Profiles and Hidden Patterns in Data
ACTION ITEMS

Researchers and business people have a powerful new tool to quickly and easily identify groups and segments in their data. Fueled by recent, major initiatives in data mining, AnswerTree, developed by SPSS Inc., enables users to easily find segments, build profiles, predict outcomes and discover patterns in data. Automatically producing an intuitive tree diagram, this multi-method classification tool provides a visual snapshot of patterns and segments. AnswerTree is now shipping and is priced at $995 in North America.

AnswerTree contains four powerful decision tree algorithms, each working best with certain types of data so users can get the best model for each situation. With a dynamic, interactive "tree" diagram, users can quickly identify and examine key groups in their data. Users get presentation quality charts and tables with multiple display options to customize their reports. AnswerTree also features many advanced features, making it easy for users to put the model to work for them.

"With AnswerTree, users can make the most of their data," said SPSS CEO and President Jack Noonan. "No other classification tool has as many tree-based algorithms, enabling users to make more informed decisions and get the best model for their data."

AnswerTree is designed for those who need to identify key groups in their data, for example, credit risk scoring, database marketing, institutional research and crime analysis. AnswerTree helps database marketers build profiles of key customers while users involved in direct mailings can easily identify who is most likely to respond to a promotion.

With the intuitive tree diagrams, quality improvement managers can quickly pinpoint the most important causes of defect. Among AnswerTree's features:

• Four algorithms. Users can build better models by choosing the algorithm best suited for their data. Choose from four powerful modeling algorithms including CHAID, Exhaustive CHAID, QUEST and Classification and Regression Trees.
• Intuitive tree diagrams. Users can identify and evaluate groups within their data. A tree diagram, charts and tables give users an instant snapshot of their segments.
• Get specific answers from easily tailored models. Users can make changes to their model quickly and easily by using the tree diagram. Branches can be pruned to reduce the number of segments. Users can combine segments by dragging and dropping tree nodes. To control the size and scope of the model, users can set the smallest segment size allowable, break-even cost and other parameters.
• Reliable results. With the partitioning and validation feature, users can train their tree model on a subset of their data and then apply it to the rest of the data for validation. Also, gains tables and mis-classification charts enable users to see incremental benefits and costs of each group. Users save time by reusing their model. Users can save the decision rules which apply to each segment in SQL code to apply the rules to a data source, for example, to generate mailing lists based upon the observed criteria.

System Requirements

• Windows 95 or Windows NT 3.51/4.0
  • -- 486DX or higher IBM-compatible PC
  • -- 40 MB hard disk space (10 MB for software; 30 MB recommended for data)
  • -- 12 MB RAM required (16 MB strongly recommended)
  • -- VGA monitor (SVGA recommended)

More information is available on the World Wide Web at http://www.spss.com