Abstract
A model for learning morphological transformations is given. The model is based upon Koskenniemi's Two Level Morphology model. It is shown that while the number of examples required for learning is polynomial, the computational problem associated with learning is intractable.
Learning homomorphisms is a simple special case of the general model. While a general method for learning homomorphisms is not known, it is shown that even if the target language is over a single letter alphabet, learning within the minimum number of examples is computationally intractable. If, however, additional examples are given, learning can be achieved in time polynomial in the size of the morphological transformation.
This is a preview of subscription content, log in via an institution to check access.
Preview
Unable to display preview. Download preview PDF.
Similar content being viewed by others
References
Angluin, D. and C. Smith, “Inductive inference: theory and methods”. Comput. Surveys, 15, 237–269, (1983)
Barton, G. E., “Computational complexity in two-level morphology”. ACL '86, 53–59.
Blumer A., A. Ehrenfeucht, D. Haussler and M. Warmuth, Learnability and the Vapnik-Chervonenkis dimension, J. ACM, 36(4), 929–965, (1989).
Harrison, M. A., Introduction to Formal Language Theory, Addison-Wesley, 1978.
M.R. Garey and D.S. Johnson, Computers and Intractability, A Guide to the Theory of NP-Completeness, W.H. Freeman and Company, (1979).
Gold, E. M., “Complexity of automaton identification from given data”. Information and Control, 37, 302–320. (1978).
Koskenniemi, K. and K. W. Church. “Complexity, Two-level morphology, and Finnish”. COLING '88.
Kataja, L., and K. Koskenniemi, “Finite-state description of Semitic morphology: A case study of Ancient Akkadian”.
Kearns, M., and L. Valiant, “Cryptographic limitations on learning Boolean formulae and finite automata”. Proc. 21st STPC, 433–444, (1989).
Koskenniemi, K., “Two-level morphology: A general computational model for word-form recognition and production”. Publication No. 11, University of Helsinki, Dept. of General Linguistics, Hallituskata 11-33, SF-00100 Helsinki 10, Finland.
A. Itai, A. Lavie, U. Ornan and M. Rimon, “On the applicability of Two Level morphology to the inflection of Hebrew verbs”. ALLC June 1988, Jerusalem, (TR 513, CS Department Technion).
Littlestone, N., “Learning quickly when irrelevant attributes abound: a new linear threshold algorithm”. Machine Learning 2, 285–318, (1987).
Littlestone, N., “Mistake bounds and logarithmic linear-threshold learning algorithms”, Ph.D. Thesis, U.C. Santa Cruz, March 1989.
Oncina, J., P.García, and E. Vidal, Learning subsequential transducers for pattern recognition interpretation tasks”. IEEE Trans. Pattern Anal. and Machine Intell., 15, 448–458, (1993)
Pitt, L., and M. K. Warmuth, “Reduction among prediction problems: on the difficulty of predicting automata”. Proc. of 3rd annual structure in complexity theory, (1988).
Valiant L.G., A Theory of the Learnable, Comm. ACM, 27(11), 1134–42, (1984).
Author information
Authors and Affiliations
Editor information
Rights and permissions
Copyright information
© 1994 Springer-Verlag Berlin Heidelberg
About this paper
Cite this paper
Itai, A. (1994). Learning morphology — practice makes good. In: Carrasco, R.C., Oncina, J. (eds) Grammatical Inference and Applications. ICGI 1994. Lecture Notes in Computer Science, vol 862. Springer, Berlin, Heidelberg. https://doi.org/10.1007/3-540-58473-0_132
Download citation
DOI: https://doi.org/10.1007/3-540-58473-0_132
Published:
Publisher Name: Springer, Berlin, Heidelberg
Print ISBN: 978-3-540-58473-5
Online ISBN: 978-3-540-48985-6
eBook Packages: Springer Book Archive