Can you recommend an automated document/image classifier for Linux? One that would take a whole mess of PDF, txt, DOCS, ODT and image files and tag them with topics, sort by dates, topics etc?