Approximate ILP Rules by Backpropagation Neural
Network: A Result on Thai Character Recognition

Boonserm Kijsirikul and Sukree Sinthupinyo

Department of Computer Engineering, Chulalongkom University,
Phayathai Rd., Phathumwan, Bangkok, 10330, Thailand
Email: Boonserm@cp.eng.chula.ac.th



Abstract. This paper presents an application of Inductive Logic Programming
(ILP) and Backpropagation Neural Network (BNN) to the problem of Thai
character recognition. In such a learning problem, there exist several different
classes of examples; there are 77 different Thai characters. Using examples
constructed from character images, ILP learns 77 rules each of which defines
each character. However, some unseen character images, especially the noisy
images, may not exactly match any learned rule, i.e., they may not be covered
by any rule. Therefore, a method for approximating the rule that best matches
the unseen data is needed. Here we employ BNN for finding such rules.
Experimental results on noisy data show that the accuracy of rules learned by
ILP without the help of BNN is comparable to other methods. Furthermore,
combining BNN with ILP yields the significant improvement and surpasses the
other methods tested in our experiment.
References

[1]	C.Baroglio and M.Botta. Multiple Predicate Learning with RTh. Topic in
Artificial Intelligence, LNAI 992, 1995.
[2]	H.Blockeel and L.De Raedt. Experiments with Top-Down Induction of Logical
Decision Trees. Technical Report CW247, Dept. of Computer Science,
K.U.Leuven, 1997.
[3]	M.Botta, A.Giordana and R.Piola. FONN: Combining First Order Logic with
Connectionist Learning. Proceedings of the 14th International Conference on
Machine Learning, 1997.
[4]	S.Dzeroski, N.Jacobs, M.Molina, C.Moure, S.Muggleton and W.V.Laer.
Detecting Traffic Problems with ILP. Proceedings of the 8th International
Workshop on Inductive Logic Programming, 1998.
[5]	S.Dzeroski, S.Schulze-Kremer, K.R.Heidtke, K.Siems and D.Wettschereck.
Applying ILP to Diterpene Structure Elucidation from 13C NMR Spectra.
Proceedings of 6th International Workshop on Inductive Logic Programming,
1996.
[6]	S.Eupiboon. Thai Handwritten Character Recognition by Considering the
Character Head. Master Thesis, KMIT, Thailand, 1988 (in Thai).
[7]	W.V.Laer, S.Dzeroski and L.D.Raedt. Multi-Class Problems and Discretization
in ICL Extended Abstract. Proceedings of the Mlnet Familiarization Workshop
on Data Mining with Inductive Logic Programming, pp.53-60, 1996.
[8]	C.Lursinsap. Speeding Up Handwritten Character Neural Learning by Surface
Projection. Proceedings of the 2nd Symposium on Natural Longuage Processing,
pp. 198-205, 1995.
[9]	L.Martin and C.Vrain. MULT_ICN: An Empirical Multiple Predicate Learner.
Proceedings of the 5th International Workshop on Inductive Logic Programming,
1995.
[10]	S.Muggleton. Inverse Entailment and PROGOL. New Generation Computing,
3:245-286, 1995.
[11]	S.Muggleton, D.Page and A.Srinivasan. An Initial Experiment into
Stereochemistry-Based Drug Design Using Inductive Logic Programming.
Proceedings of the 6th International Workshop on Inductive Logic Programming,
pp.25-40, 1996.
[12]	P.Phokharatkul and C.Kimpan. Recognition of Handprinted Thai Characters
Using the Cavity Features of Character Based on Neural Network. Proceedings
of the 1998 IEEE Asia-Pacific Conference on Circuits and Systems, pp. 149-152.
1998.
[13]	D. Rattanatarn and S. Jitapunkul. Thai Printed Characters Recognition Using the
Fuzzy Logic Technique and Syntactic Method. Proceedings of the 18th
Conference of Electrical Engineering, 1995 (in Thai)
[14]	D.E.Rumelhart, G.E.Hinton and R.J.Williams. Learning internal representations
by error propagation. In D. E. Rumelhart, & J. L. McClelland (Eds.), Parallel
distributed processing (Vol 1), Cambridge, MA:MIT Press, 1986.
[15]	T.Tanprasert and C.Tanprasert, Variable Simulated Light Sensitive Model for
Handwritten Thai Digit Recognition. Proceedings of the 2nd Symposium on
Natural Language Processing, pp. 173-179, 1995.
[16]	G.G.Towell and J.W.Shavlik. Knowledge-Based Artificial Neural Networks.
Artificial Intelligence, 70 (4):119-116, 1994.
