Frame Problem

The frame problem confronts classical systems that operate by creating and updating internal representations (Ford & Pylyshyn, 1996; Pylyshyn, 1987).  The frame problem faces any system that has to update classical descriptions of a changing world.  In order to represent a situation after some action has occurred, or after a property in the world has changed for some other reason, a system is not only faced with deducing what changes are produced by the action, but also must deduce what changes are not produced by the action.  “In fact, very little of what from a purely logical point of view could happen as a result of an action, does happen. [...] The outcome is a problem solver sitting in the midst of a sea of axioms representing non-changes, endlessly calculating non-effects” (Janlert, 1987, p. 6).  One solution to the frame problem is to artificially limit the depth of deductions to be considered.  This can be accomplished, for instance, by restricting processes to be modular instead of isotropic (Fodor, 1983).

References:

1. Fodor, J. A. (1983). The Modularity Of Mind. Cambridge, MA: MIT Press.
2. Ford, K. M., & Pylyshyn, Z. W. (1996). The Robot's Dilemma Revisited: The Frame Problem In Artificial Intelligence. Norwood, N.J.: Ablex Pub.
3. Janlert, L.-E. (1987).  Modeling change -- The frame problem.  In Z.W. Pylyshyn (Ed.) The Robot's Dilemma.  Norwood, NJ: Ablex Publishing.
4. Pylyshyn, Z. W. (1987). The Robot's Dilemma: The Frame Problem In Artificial Intelligence. Norwood, NJ: Ablex.

(Added March 2011)