Theory formation in artificial intelligence

ABSTRACT

Theories in terms of causal mechanisms and causal relationships are a critical component of problem solving in artificial intelligence. A theory for explaining a given observation should satisfy constraints based on causal knowledge. In this paper, we present a new approach to theory formation. Under this approach, a theory is formed by reasoning with causal constraints. The reasoning method is constraint-satisfaction. Each coherent set of causal mechanisms discovered by the method instantiates the domain causal model to generate a causal hypothesis. If the domain causal model is true, then it can be shown that one of the causal hypotheses generated is true. In the case of using multi-level constraints, a theory is refined into more details by reasoning top-down through the levels of constraints.

REFERENCES

• 1: Bauchanan , B.G. and Shotliffe, E.h. Rule based expert Systems, Addison-Wesley Publishing Company, Inc., Reading Massachusetts, 1984.
• 2: R. Dechter , J. Pearl, Network-based heuristics for constraint-satisfaction problems, Artificial Intelligence, v.34 n.1, p.1-38, Dec. 1987 [doi>10.1016/0004-3702(87)90002-6]
• 3: J de Kleer , B C Williams, Diagnosing multiple faults, Artificial Intelligence, v.32 n.1, p.97-130, April 1987 [doi>10.1016/0004-3702(87)90063-4]
• 4: Doyle, R.J., Constructing and refining causal explanations from an inconsistent domain theory, in Proceedings of AAAI-86, Philadelphia, 1986, 538-544.
• 5: B J Kuipers, Qualitative simulation, Artificial Intelligence, v.29 n.3, p.289-338, Sept. 1986 [doi>10.1016/0004-3702(86)90073-1]
• 6: Tom M. Mitchell , Richard M. Keller , Smadar T. Kedar-Cabelli, Explanation-Based Generalization: A Unifying View, Machine Learning, v.1 n.1, p.47-80 [doi>10.1023/A:1022691120807]
• 7: Patil, R.S., Szolovits, P., and Schwartz, W.B., Causal understanding of patient illness in medical diagnosis, 7th IJCAI, 1981, 893-899.
• 8: Ramesh S. Patil, A case study on evolution of system building expertise: medical diagnosis, AI in the 1980s and beyond, Massachusetts Institute of Technology, Cambridge, MA, 1990
• 9: Pazzani, M, Dyer, M., and Flowers, M., The role of prior causal theories in generalization, in Proceedings of AAAI-86, Philadelphia, 1986, 545-550.
• 10: Judea Pearl , Michael Tarsi, Structuring causal trees, Journal of Complexity, v.2 n.1, p.60-77, March 1986 [doi>10.1016/0885-064X(86)90023-3]
• 11: J Pearl, Fusion, propagation, and structuring in belief networks, Artificial Intelligence, v.29 n.3, p.241-288, Sept. 1986 [doi>10.1016/0004-3702(86)90072-X]
• 12: J Pearl, Evidential reasoning using stochastic simulation of causal models, Artificial Intelligence, v.32 n.2, p.245-257, May 1987 [doi>10.1016/0004-3702(87)90012-9]

INDEX TERMS

Primary Classification:
  I. Computing Methodologies
  I.2 ARTIFICIAL INTELLIGENCE

Additional Classification:
  I. Computing Methodologies
  I.2 ARTIFICIAL INTELLIGENCE
      I.2.8 Problem Solving, Control Methods, and Search
          Subjects: Heuristic methods

General Terms:
Theory