Searches Improve As Systems Learn From QueriesSearches Improve As Systems Learn From Queries

Over the past few years, customers of customer-relationship management vendor Kana Inc. told the company in growing numbers that they want to expose their support knowledge base to customers via the Web. Kana has responded with tools to make that happen, but the utilities don't always make connections between relevant data.

The release this week of search-intelligence vendor Verity Inc.'s K2 Developer 5.0 suite is the first step toward making those connections happen automatically. A developer's kit in the K2 suite, due later this year, features a set of APIs designed to let software makers build automatic content-organization and classification capabilities into their applications.

By integrating those capabilities into its CRM platform, which it plans to do by year's end, Kana lets its customers benefit from more intelligent content categorization while reducing costs related to the building of taxonomies and metadata. "They will no longer have to have as many knowledge administrators managing and monitoring their knowledge bases," says H.A. Schade, director of product management at Kana. Users accessing a Kana-built self-service Web application using the Verity engine will get better answers to questions as the system learns how to expand its horizons in responding to natural-language searches.