Contact: Estelle Brand
Initially released in June 2000, XAffinity® was one of the first data mining products to be database enabled, and goes farther than most in-place mining tools in terms of database integration. Not only is mining done directly from database tables or views, but SQL is used to analyze the data and place the mining results back into the database, allowing analysts to browse the rules with standard report and OLAP tools.
XAffinity® was designed to provide exceptional flexibility for association and sequence pattern detection and analysis. The initial release included innovative extensions to classic associations and sequencing. Key among these are automatic partitioning to allow for easy generation of rules across segments such as day of week or customer type and the rule discriminator which assists in identifying the rules that vary the most, by lift, confidence or support, across the partitioning variable.
XAffinity® has always supported multiple time period dependencies – later, next visit, and interval – as well as complete interval flexibility “making it suitable for use in unique applications such as Anticipatory Maintenance and click-stream analysis” said Rob Gerritsen, President of Exclusive Ore.
XAffinity® 4.0 continues to build on this solid foundation, with key new features that expand the data preparation features and the information included in sequence rules.
Ø XAStatistics. Designed to facilitate data preparation and understanding, XA statistics produces summary statistics for all columns in a table, across all rows viewable either as text or graphics. Consistent with XAffinity’s powerful partitioning feature, one column can optionally be chosen as a partitioning column, to enable cross-tabulated against the partitioning column. Resulting statistics can be displayed as 2D or 3D histograms, line charts or pies. Cross tabs can quickly be pivoted, and 3D displays can be zoomed and rotated.
Ø Enhanced sequence rules for sequencing: XAffinity 4.0 expands sequence rules to include interval (or lag) statistics, including shortest, longest and mean interval for each paired set of events. For example, in retail, a sequence rule that relates product the purchase of outdoor dining furniture to a subsequent purchase of a barbeque will also indicate the shortest time and longest time between such sequential purchases as well as the average time. This helps marketers plan the appropriate time for cross sells suggestions.
Interval statistics are particularly useful when using XAffinity for Anticipatory Maintenance applications. Such statistics can be used to differentiate urgent repairs (those with a small minimum delay) from repairs that can be delayed until a subsequent maintenance period
Ø Integration with MS Office through drag and drop. Selected rules from any analysis can be exported to Excel or Word through copy and paste. An entire rule set can also be exported in column delimited format to enable further or off-line analysis.
XAffinity® can be installed on computers running Microsoft Windows NT 4.0, Windows 2000 or Windows XP, and can be integrated with a number of database products, including Microsoft SQL Server, Microsoft Access, Oracle, IBM DB2, and Red Brick.
XAffinity® is available both as an end-user application for marketers and web analysts, and as ActiveX DLL for building turn-key applications. For additional information about XAffinity® or to receive a 30-day free trial, please visit the Exclusive Ore web site at www.exclusiveore.com.
Exclusive Ore Inc. was founded in 1997 as a consulting and software services company whose sole emphasis was the delivery of quality data mining solutions. Since that time, Exclusive Ore has successfully implemented solutions in a wide variety of industries including Telco, Retail, Finance, and Consumer Packaged Goods.