OCR-ed text as ‘tag’ for a scanned document helps identify the document by its content.