Open-Source Search

IBM plans to make its enterprise search middleware, designed to facilitate searches of unstructured data, available as open-source code. It's called Unstructured Information Management Architecture, and IBM says more than 15 knowledge-management companies intend to support it as a standard framework.

The company also is wedding the latest iteration of its WebSphere Information Integrator OmniFind Edition with UIMA. Its goal is to make enterprise search results more relevant and make it easier to apply third-party analytics software.

UIMA is a software infrastructure layer that supports the search and analysis of disorganized data. Unstructured data--E-mail messages, Word documents, and the like--isn't easily classified, sorted, or searched.

The open-source move could let software partners add value to unstructured content with analytics applications, says Dana Gardner, principal analyst with consulting firm Interarbor Solutions.

One example of the software that's being developed to take advantage of UIMA is called Quality Early Warning, an app that uses text analytics technology from ClearForest Corp. and Cognos Inc. to help identify quality issues in the auto industry. It identifies problems by finding, understanding, and correlating data culled from warranty claims, maintenance records, repair requests, and call-center logs. That lets companies take quicker corrective action.