Can a computer laugh? Can a robot laugh? More importantly, is it possible for a machine to "get" a joke so they know what to laugh at? Igor M. Suslov of the Kapitza Institute for Physical Problems in Moscow, Russia has published a paper online titled, "Can a Computer Laugh" (PDF format). In it he discusses theoretical reasons for laughter and how the ideas can be applied in data processing. In human cognition, humor appears to correspond to specific types of information processing "malfunctions" that need to be flagged to prevent incorrect learning. In effect, laughing corresponds to quick nullification in some types of neural networks. The author offers a computer model that can recognize the humor in specific technical types of jokes identified by Freud.