American Heart Association needed to migrate several heterogeneous and highly non-standard legacy internets and intranets to a new content management system which was to support internal and external portals.

The legacy content consisted of 6,000 pages of heterogeneous and highly non-standard HTML. The content needed to be disassembled into discrete components to be managed separately in a content database. A change in requirements resulted in the need to reinsert images that had been removed in an earlier phase of content conversion.

Intelligent Systems applied pattern recognition techniques to automatically extract "body" content from legacy pages and insert the content into a content database. A hybrid process combining automated and manual conversion was adopted. The pattern recognition tools were adapted to support the insertion of images to the appropriate location within the text.