EMC did not release product details but indicated that it would ship in the fourth quarter, according to George Symons, chief technology officer for information management at the Hopkinton, Mass.-based company.

Initially, the technology will focus on unstructured files such as text files, spreadsheets, PowerPoint presentations and semistructured files such as e-mail. It will eventually support databases as well, Symons said. It will enable administrators to set up four to 10 classes to which data could be assigned, along with retention requirements, who has access to it, and compliance requirements, he said.

Such a system could also be set up to automatically delete information based on criteria such as how much time has passed since it was accessed, Symons said. Setting up policies and processes for such deletion will help organizations follow compliance and e-discovery requirements by demonstrating that they have a policy and a process and that files are not being deleted randomly, he said.

Brian Babineau, an analyst at Enterprise Strategy Group in Milford, Mass., said EMC has all the pieces to create a data-classification product but must now knit them together.

Though EMC is using technology from prior acquisitions, such as Legato Systems Inc., Documentum Inc. and Smarts Inc., to develop the new data-classification technology, the company is not acquiring any of the data-classification start-ups to jump-start development, said Symons. This is actually unusual for EMC because the company has a history of producing innovation through acquisition and is typically a bellwether in the storage industry, said Simon Robinson, an analyst at The 451 Group Inc. in New York.