Detecting content drift on the web using web archives and textual similarity