This Article 
 Bibliographic References 
 Add to: 
ENIGMA: A System That Learns Diagnostic Knowledge
February 1993 (vol. 5 no. 1)
pp. 15-28

The results of extensive experimentation aimed at assessing the concrete possibilities of automatically building a diagnostic expert system, to be used in-field in an industrial domain, by means of machine learning techniques, are described. The system, ENIGMA, is an incremental version of the ML-SMART system, which acquires a network of first-order logic rules, starting from a set of classified examples and a domain theory. An application is described that consists of discovering malfunctions in electromechanical apparatus. ENIGMA's efficacy in acquiring sophisticated knowledge and handling complex structured examples is largely due to its underlying database management system, which supports the learning operators, defined at the abstract level, with a set of primitives, taken from the field of deductive databases. An expert system, MEPS, devoted to the same task, has also been manually developed. A number of comparisons along different dimensions of the manual and automatic development process have been possible, allowing some practical indications to be suggested.

[1] R. S. Michalski and R. L. Chilauski, "Learning by being told and learning from examples: An experimental comparison of the two methods of knowledge acquisition,"Int. J. Policy Analysis Inform. Syst., vol. 4, pp. 125-126, 1980.
[2] R. Michalski, I. Mozetic, J. Hong, and N. Lavrac, "The multi-purpose incremental learning system AQ15 and its testing application to three medical domains," inProc. Fifth Natl. Conf. on Artificial Intelligence, Philadelphia, PA, 1986, pp. 1041-1045.
[3] B. Buchanan and T. Mitchell, "Model-driven learning of production rules," in D. Waterman and F. Hayes-Roth, eds.,Pattern-Directed Inference Systems. New York: Academic, 1978, pp. 297-312.
[4] S. Minton, J. Carbonell, C. Knoblock, D. Kuokka, O. Etzioni, and Y. Gil, "Explanation based learning: A problem solving perspective,"Artificial Intell., vol. 40, pp. 63-118, 1989.
[5] J. R. Quinlan, "Induction of decision trees,"Machine Learning, vol. 1, no. 1, pp. 81-106, 1986.
[6] B. Cestnik, I. Kononenko, and I. Bratko, "ASSISTANT-86: A knowledge elicitation tool for sophisticated users," in I. Bratko and N. Lavrac, Eds.,Progress in Machine Learning. Wilmslow, U.K.: Sigma, 1987, pp. 31-45.
[7] F. Bergadano, A. Giordana, and L. Saitta, "Automated concept acquisition in noisy environments,"IEEE Trans. Pattern Anal. Mach. Intell., vol. 10, pp. 555-578, 1988.
[8] P. H. Winston, "Learning structural descriptions from examples," inThe Psychology of Computer Vision. New York: McGraw Hill, 1975, pp. 157-209.
[9] S. A. Vere, "Induction of concepts in the predicate calculus," inProc. IJCAI-75, Tbilisi, USSR, 1975, pp. 351-356.
[10] J.R. Quinlan, "Learning Logical Definitions from Relations,"Machine Learning, Vol. 5, No. 3, Aug. 1990, pp. 239-266.
[11] S. Muggleton and C. Feng, "Efficient induction of logic programs," inProc. First Conf. on Algorithmic Learning Theory, Tokyo, Japan, 1990.
[12] R. Gemello, F. Mana, and L. Saitta, "Rigel: An inductive learning system,"Mach. Learning J., vol. 6, pp. 7-36, 1991.
[13] F. Hayes-Roth and J. McDermott, "An interference matching technique for inducing abstractions,"Commun. ACM, vol. 21, no. 5, pp. 401-410, 1978.
[14] W. Iba, J. Wogulis, and P. Langley, "Trading off simplicity and coverage in incremental concept learning," inProc. Fifth Int. Conf. on Machine Learning, Ann Arbor, MI, 1988, pp. 73-79.
[15] Y. Kodratoff, "Improving the generalization step in learning," in R. Michalski, J. Carbonell, and T. Mitchell, eds.Machine Learning: An Artificial Intelligence Approach. Palo Alto, CA: Morgan Kaufmann, 1986, pp. 215-244.
[16] T. M. Mitchell, R. H. Keller, and S. T. Kedar-Cabelli, "Explanation-based generalization: A unifying view,"Machine Learning, vol. 1, no. 1, pp. 47-80, 1986.
[17] G. DeJong and R. Mooney, "Explanation based learning: An alternative views,"Machine Learning, vol. 1, no. 2, pp. 145-176, 1986.
[18] F. Bergadano, R. Gemello, A. Giordana, and L. Saitta, "SMART: A problem solver for learning from examples,"Fundamenta Inform., vol. 12, pp. 29-50, 1989.
[19] F. Bergadano and A. Giordana, "A knowledge intensive approach to concept induction," inProc. Fifth Int. Conf. on Machine Learning, Ann Arbor, MI, 1988, pp. 305-317.
[20] F. Bergadano, A. Giordana, and S. Ponsero, "Deduction in top-down inductive learning," inProc. 6th Int. Workshop on Machine LearningIthaca, NY, 1989, pp. 23-25.
[21] F. Bergadano, A. Giordana, L. Saitta, D. De Marchi, and F. Brancadori, "Integrated learning in a real domain," inProc. 7th Int. Conf. on Machine Learning, Austin, TX, 1990, pp. 322-329.
[22] F. Bergadano, A. Giordana, and L. Saitta, "Biasing induction by using a domain theory: An experimental evaluation," inProc. ECAI-90, Stockholm, Sweden, 1990, pp. 84-89.
[23] F. Bergadano, A. Giordana, and L. Saitta, "Learning from examples in presence of uncertainty," in E. Sanchez and L. Zadeh, eds.Approximate Reasoning in Intelligent Systems, Decision and Control. Headington, U.K.: Pergamon, 1987, pp. 105-124.
[24] R. S. Michalski, "A theory and methodology of inductive learning,"Artif. Intell., vol. 20, pp. 111-161, 1983.
[25] L. Vieille, "Recursive axioms in deductive databases: The query sub-query approach," inProc. First Int. Conf. on Expert Database Systems, Charleston, SC, 1986.
[26] J. D. Ullman,Principles of Databases Systems. Rockville, MD: Computer Science Press, 1982.
[27] L. J. Henschen and S. Naqvi, "On compiling queries in recursive first-order databases,"J. ACM, vol. 1, no. 1, pp. 47-85, 1984.
[28] G. Gardarin and E. Simon, "Les systèmes de gestion de bases de données déductives,"Synthèse, vol. 6, 1987.
[29] M. Botta and A. Giordana: "Learning quantitative features in a symbolic environment,"Lecture Notes in Artif. Intell., vol. 542, pp. 296-305, 1991.
[30] F. Brancadori and S. Radicchi, "Acquisizione automatica della conoscenza,"Inform. Oggi, vol. 9, pp. 81-93, 1989.
[31] F. Bergadano and A. Giordana, "Pattern classification: An approximate reasoning framework,"Int. J. Intell. Syst., vol. 4, pp. 23-44, 1989.
[32] M. Botta, S. Ravotto, and L. Saitta: "Use of causal models and abduction in learning diagnostic knowledge,"Int. J. Man-Machine Studies, vol. 6, pp. 289-307, 1992.
[33] L. Saitta, M. Botta, S. Ravotto, and S. Sperotto: "Improving learning by using deep models," inProc. First Int. Workshop on Multistrategy Learning, Harpers Ferry, WV, 1991, pp. 131-143.
[34] M. Botta, A. Giordana, and L. Saitta, "Learning in uncertain environments," in R. Yager and L. Zadeh, Eds.,An Introduction toFuzzy Logic Applications in Intelligent Systems. Boston, MA: Kluwer Academic, 1992, pp. 281-296.

Index Terms:
diagnostic knowledge learning; ENIGMA; diagnostic expert system; machine learning; ML-SMART system; first-order logic rules; domain theory; database management system; deductive databases; MEPS; deductive databases; diagnostic expert systems; learning (artificial intelligence)
A. Giordana, L. Saitta, F. Bergadano, F. Brancadori, D. De Marchi, "ENIGMA: A System That Learns Diagnostic Knowledge," IEEE Transactions on Knowledge and Data Engineering, vol. 5, no. 1, pp. 15-28, Feb. 1993, doi:10.1109/69.204088
Usage of this product signifies your acceptance of the Terms of Use.