Macro-operators: A weak method for learning (Q1057073)

The author gathers already known ideas used in problem solving and develops a nice united theory of a weak method for learning; this method is based on macro-operators (sequences of primitive operators which perform a subgoal so that the previous obtained subgoals are preserved) and on the property of serial decomposability of primitive operators; the method successfully solves a class of problems unsatisfactorily treated so far, i.e. problems with non-serializable subgoals; also the author describes and analyses the performance of both, a problem-solving program and a learning program for the aforementioned method.