A system for processing text captured from rendered documents is
described. The system receives a sequence of one or more words optically
or acoustically captured from a rendered document by a user. The system
identifies among words of the sequence a word with which an action has
been associated. The system then performs the associated action with
respect to the user.