SPSS Inc. this week debuted a release of its Clementine predictive-analytics application with new capabilities that aid in fraud-detection, customer-analysis, and revenue-assurance applications.
Clementine 10 offers a new anomaly-detection algorithm that identifies data that doesn't fit an established pattern or signifies an unusual event. The algorithm simplifies analysis and scoring and helps with such tasks as fraud detection and spotting financial inaccuracies.
The new feature-selection capability will help developers produce predictive models more easily and determine what data attributes best fit a model and are relevant to a specific problem. SPSS says such capabilities are particularly applicable to customer-relationship-management analysis tasks such as customer acquisition, customer retention, and upsell/cross-sell analysis.
Test and measurement device manufacturer National Instruments Corp. uses Clementine for a number of forecasting applications, including placing values on sales prospects. Dan Courtright, a market analyst at the company, has been beta testing Clementine 10, including using it to help identify data-quality problems. He's particularly pleased with its improved ability to import and export data to and from third-party data sources.
The software also provides a slew of data-preparation and productivity enhancements. Clementine 10 is available now, priced at $80,000 per server and $5,000 per user.