Prolog, Refinements and RLGGs

Claude Sammut

School of Computer Science and Engineering
University of New South Wales
Sydney 2052 Australia
claude@cse.unsv.edu.au
http://vvv.cse.unsw.edu.au/~claude



Abstract. Cohens [1] refinement rules provide a flexible mechanism
for introducing intentional background knowledge in an ILP system.
Whereas Cohen used a limited second order theorem prover to implement 
the rule interpreter, we extend the method to use a full Prolog
interpreter. This makes the introduction of more complex background
knowledge possible. Although refinement rules have been used to generate 
literals for a general-to-specific search, we show how they can also be
used as filters to reduce the number of literals in an RLGG algorithm.
Each literal constructed by the LGG is tested against the refinement
rules and only admitted if a refinement rule has been satisfied.
References

1.	Cohen, W.: Learning to classify English text with ILP methods. In L. de Raedt
(Eds.), Advances in Inductive Logic programming. IOS Press (1996) 124143
2.	Muggleton, S., Feng, C.: Efficient induction of logic programs. In First Conference
on Algorithmic Learning Theory. Omsha, Tokyo (1990)
3.	Muggleton, S.: Inverse Entailment and Progol. New Generation Computing 13
(1995) 245286.
4.	Page, C. D., Frisch, A. M.: Generalization and Learnability: A study of constrained
atoms. In S. Muggleton (Eds.): Inductive Logic Programming. Academic Press
(1992) 2961
5.	Plotkin, G. D.: A further note on inductive generalization. In B. Meltzer and D.
Michie (Eds.): Machine Intelligence 6. Elsevier, New York (1971)
6.	Quinlan, J. R.: Learning Logical Definitions from Relations. Machine Learning 5
(1990) 239-266
7.	Rouveirol, C., Puget, J.-F.: Beyond Inversion of Resolution. In Proceedings of the
Seventh International Conference on Machine Learning. Morgan Kaufmann (1990)
8.	Sammut, C. A., Banerji, R. B.: Learning Concepts by Asking Questions. In R.
S.	Michalski Carbonell, J.G. and Mitchell, T.M. (Eds.): Machine Learning: An
Artificial Intelligence Approach, Vol 2. Morgan Kaufmann, Los Altos, California
(1986) 167192
9.	Sammut, C.: Using background knowledge to build multistrategy learners. Machine
Learning 27 (1997) 241257
10.	Srinivasan, A., Camacho, R.: Experiments in numerical reasoning with Inductive
Logic Programming. Journal of Logic Programming (in press)
11.	Zrimec, T., Sammut, C.A.: A Medical Image Understanding System. Engineering
applications of Artificial Intelligence 10(1) (1997) 3139.
