Abstract
In this paper we study induction of unrestricted clausal theories from interpretations. First, we show that in the propositional case induction from complete evidence can be seen as an equivalence-preserving transformation from DNF to CNF. From this we conclude that induction is essentially a process of determining what is false in the domain of discourse. We then proceed by investigating dual normal forms for evidence and hypotheses in predicate logic. We define evidence nonnal form (ENF), which is Skolemised existential DNF under a Consistent Naming Assumption. Because ENF is incomplete, in the sense that it does not have the expressive power of clausal logic, ENF evidence requires the identification of Skolem terms. The approach is partly implemented in the Primus system.
Part of this work was supported by Esprit IV Long Term Research Project 20237 (ILP II). Thanks are due to Luc De Raedt for inspiring discussions and useful suggestions.
Preview
Unable to display preview. Download preview PDF.
Similar content being viewed by others
References
D. Angluin, M. Frazier & L. Pitt. Learning conjunctions of Horn clauses. Machine Learning, 9(2/3):147–164, 1992.
N.H. Bshouty. Exact learning via the monotone theory. Proc. IEEE Symp. on Foundations of Computer Science, pp.302–311. 1993.
P. Clark & T. Niblett. The CN2 induction algorithm. Machine Learning, 3(4):261–283, 1989.
R. Dechter & J. Pearl. Structure identification in relational data. Artificial Intelligence, 58:237–270, 1992.
L. De Raedt. Induction in logic. Proc. 3d Multistrategy Learning Workshop, pp.29–38. 1996.
L. De Raedt & L. Dehaspe. Clausal discovery. Machine Learning, 26(2/3):99–146, 1997.
L. De Raedt & W. Van Laer. Inductive Constraint Logic. Proc. 6th Workshop on Algorithmic Learning Theory, pp.80–94. Springer, 1995.
L. De Raedt, N. Lavrac & S. Dzeroski. Multiple predicate learning. Proc. 13th Int. Joint Conf. on Artificial Intelligence, pp.1037–1042. Morgan Kaufmann, 1993.
A. Friedman. Fundamentals of Logic Design and Switching Theory. Computer Science Press, 1986.
R. Khardon & D. Roth. Reasoning with models. Technical Report TR-1-94, Center for Research in Computing Technologies, Harvard University, 1994.
H. Kautz, M. Keams & B. Selman. Horn approximations of empirical data. Artificial Intelligence, 74:129–145, 1995.
Author information
Authors and Affiliations
Editor information
Rights and permissions
Copyright information
© 1997 Springer-Verlag Berlin Heidelberg
About this paper
Cite this paper
Flach, P.A. (1997). Normal forms for inductive logic programming. In: Lavrač, N., Džeroski, S. (eds) Inductive Logic Programming. ILP 1997. Lecture Notes in Computer Science, vol 1297. Springer, Berlin, Heidelberg. https://doi.org/10.1007/3540635149_43
Download citation
DOI: https://doi.org/10.1007/3540635149_43
Published:
Publisher Name: Springer, Berlin, Heidelberg
Print ISBN: 978-3-540-63514-7
Online ISBN: 978-3-540-69587-5
eBook Packages: Springer Book Archive