Briefs

16th October 2006. Fair Isaac Corp has introduced new spidering capabilities into its modeling tool that promise to speed up development and deployment of predictive analysis applications.

The Data Spiders, which is included as an optional module in Model Builder 5.3, use genetic algorithms to scour across transactional data stores and summarize data to find and find the most effective data sets for predictive analysis.

The software works by extracting and testing predictive characteristics, like patterns, for use in models. For example, it can search across clickstreams for patterns to detect or call center records to prevent churn.

Officials at Minneapolis-based Fair Isaac said the spiders remove much of the manual programming graft in the pre-staging and preparation of data for predictive analysis and data mining applications.

Read more.