This Article 
 Bibliographic References 
 Add to: 
LEW: Learning by Watching
March 1990 (vol. 12 no. 3)
pp. 294-308

LEW (learning by watching), a machine learning system, is described. It was designed for knowledge acquisition in cooperation with an expert. LEW learns from examples of problem-solution (or question-answer) pairs by generalizing on differences in those pairs. In this sense, it belongs to the family of inductive learning methods. It provides for using background knowledge through the environment component of problem-solution pairs, thereby making constructive learning possible. The user can control the extent of the generalizations performed by LEW. The learning method is incremental and, to some extent, noise-resistant. The authors give an informal overview of the knowledge representation and the basic learning algorithm of LEW and indicate that the system's design meets the stated criteria and enables it to give helpful assistance, even in situations characterized by noisy or conflicting information and by lack of extensive background knowledge. The theory behind LEW is presented, along with rigorous definitions of its fundamental concepts and a general description of its learning algorithm. LEW's functioning with some larger examples, one from the QUIZ Advisor domain and another from the domain of block-world planning, is illustrated. The authors compare LEW with several other knowledge acquisition tools and introduce a precise characterization of learning from near misses and a near-miss metric. Possible extensions and enhancements to the system are noted.

[1] K. Akama and A. Ichikawa, "A basic model for learning systems," inProc. 6th Int. Joint Conf. Artificial Intelligence, Tokyo, 1979, pp. 4-6.
[2] J. H. Boose and J. M. Bradshaw, "Expertise transfer and complex problems," inProc. Knowledge Acquisition for Knowledge-Based Systems Workshop, Banff, Alta., Canada, 1986, pp. 241-247.
[3] J. H. Boose and B. R. Gaines, Eds.,Knowledge Acquisition for the Knowledge-Based Systems Workshop, Banff, Alta., Canada, 1986.
[4] A. Bundy, B. Silver, and D. Plummer, "An analytical comparison of some rule-learning programs,"Art. Intell., vol. 27, pp. 137-181, 1985.
[5] P. Constant, S. Matwin, and F. Oppacher, "LEW-P: Learning by watching in the planning domain," inProc. 8th Conf. Canadian Soc. Computational Studies of Intelligence, Edmonton, Alta., Canada, 1988, pp. 242-248.
[6] P. Constant, S. Matwin, and S. Szpakowicz, "Question-driven approach to the construction of knowledge-based software advisor systems," inProc. Third Conf. AI Applications, Orlando, FL, 1987, pp. 29-37.
[7] G. DeJong and R. Mooney, "Explanation based learning: An alternative views,"Machine Learning, vol. 1, no. 2, pp. 145-176, 1986.
[8] G. F. DeJong, R. Stepp, R. Mooney, P. O'Rorke, S. Rajamoney, A. Segre, and J. Shavlik, "Current research in machine learning," Univ. Illinois at Urbana-Champaign, Tech. Rep. TR 64, 1985.
[9] T. Dietterich and R. Michalski, "A comparative review of selected methods for learning from examples," inMachine Learning, vol. 1, R. S. Michalski, J. Carbonell, and T. Mitchell, Eds. Morgan-Kaufmann, 1983.
[10] D. Fisher, "Knowledge acquisition via incremental conceptual clustering,"Mach. Learning, vol. 2, pp. 139-172, 1987.
[11] G. Kahn, S. Nowlan, and J. McDermott, "Strategies for knowledge acquisition,"IEEE Trans. Pattern Anal. Machine Intell., vol. PAMI- 7, no. 5, pp. 511-530, 1985.
[12] D. Klahr, P. Langley, and R. Neches, Eds.,Production System Models of Learning and Development. Cambridge, MA: MIT Press, 1987.
[13] J. Kolodner,Retrieval and Organization Strategies in Conceptual Memory. Hillsdale, NJ: Lawrence Erlbaum, 1984.
[14] P. Langley and J. Carbonell, "Language acquisition and machine learning," Univ. California, Irvine, Tech. Rep. TR 86-12, 1986.
[15] M. Lebowitz, "RESEARCHER: An experimental intelligent information system," inProc. 9th Int. Joint Conf. Artificial Intelligence, Los Angeles, CA, 1985, pp. 858-862.
[16] M. Lebowitz, "Experiments with incremental concept formation: UNIMEN,"Mach. Learning, vol. 2, pp. 103-138, 1987.
[17] S. Matwin and F. Oppacher, "Knowledge acquisition by incremental learning from problem-solution pairs,"Computat. Intell., 1989.
[18] R. S. Michalski, "A theory and methodology of inductive learning," inMachine Learning, vol. 1, R. S. Michalski, J. Carbonell, and T. Mitchell, Eds. Morgan-Kaufmann, 1983, pp. 83-134.
[19] R. S. Michalski, J. Carbonell, and T. Mitchell, Eds.,Machine Learning, vol. 2, Morgan-Kaufmann, 1986.
[20] R. S. Michalski and R. L. Chilausky, "Learning by being told and learning from examples,"Int. J. Policy Anal. Inform. Syst., vol. 4, no. 2, 1980.
[21] R. S. Michalski and L. Watanabe, "Constructive closed loop learning," AI Center, George Mason Univ., 1988.
[22] T. M. Mitchell, J. C. Carbonell, and R. S. Michalski, Eds.,Machine Learning: A Guide to Current Research. Amsterdam, The Netherlands: Kluwer, 1985.
[23] 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.
[24] T. Mitchell, S. Mahadevan, and L. Steinberg, "LEAP: A learning apprentice for VLSI design," inProc. 9th Int. Conf. Artificial Intelligence, Los Angeles, CA, 1985, pp. 573-580.
[25] T. Mitchell, P. Utgoff, and R. Banerji, "Learning by experimentation," inMachine Learning, vol. 1, R. S. Michalski, J. Carbonell, and T. Mitchell, Eds. Morgan-Kaufmann, 1983, pp. 163-190.
[26] P.G. Politakis,Empirical Analysis for Expert Systems, Pitman Advanced Publishing, Boston. 1985.
[27] J. R. Quinlan, "The effect of noise on concept learning," inMachine Learning, vol. 2, R. S. Michalski, J. Carbonell, and T. Mitchell, Eds. Morgan-Kaufmann, 1986.
[28] R. Schank,Dynamic Memory: A Theory of Reminding and Learning in Computers and People. Cambridge UK: Cambridge Univ. Press, 1982.
[29] L. Siklossy, "Natural language by computer," inRepresentation and Meaning: Experiments with Information Processing, Simon and Siklossy, Eds. Englewood Cliffs, NJ: Prentice-Hall, 1972.
[30] S. Vere, "Induction of relational productions in the presence of background information," inProc. 5th Int. Joint Conf. Artificial Intelligence, Cambridge, MA, 1977, pp. 349-355.
[31] P. M. Winston,Artificial Intelligence. Reading, MA: Addison-Wesley, 1984.
[32] P. H. Winston, "Learning by augmenting rules and accumulating censors," inMachine Learning, vol. 2, R. S. Michalski, J. Carbonell, and T. Mitchell, Eds. Morgan-Kaufmann, 1986, pp. 45-62.
[33] J. G. Wolff, "Language acquisition, data compression, and generalization,"Lang. Commun., vol. 2, pp. 57-89, 1982.

Index Terms:
expert systems; problem-solution pairs; noise resistance; question-answer pairs; learning by watching; LEW; machine learning system; knowledge acquisition; inductive learning; background knowledge; knowledge representation; QUIZ Advisor; block-world planning; near-miss metric; knowledge acquisition; knowledge based systems; learning systems
P. Constant, S. Matwin, F. Oppacher, "LEW: Learning by Watching," IEEE Transactions on Pattern Analysis and Machine Intelligence, vol. 12, no. 3, pp. 294-308, March 1990, doi:10.1109/34.49054
Usage of this product signifies your acceptance of the Terms of Use.