Automating Problem Reformulation

Abstract

Research in the fields of problem solving, expert systems, and machine learning has been converging on the issue of problem representation. A system’s ability to solve problems, answer questions, and acquire knowledge has traditionally been bounded by its initial problem representation. One solution to this dilemma is to develop systems which can automatically alter their problem representation. This research deals with automatically shifting from one problem representation to another representation which is more efficient, with respect to a given problem solving method and a given problem class. The basic model of Sojourner is to derive shifts of representation from an analysis of the state space representation for a single training instance from a given problem class. These new representations are then used on subsequent problems in the same problem class. In this paper a brief overview is given of two types of representational shifts: deriving problem reductions and deriving iterative macro-operators.