On the Necessity of Special Mechanisms for Handlina Types in Inductive Logic Programming

摘要

This paper demonstrates the necessity of spe- cial handling mechanisms for type (or sort) information when learning logic programs on the basis of background knowledge that includes type hierarchy. We have developed a novel re- lational learner RHB, which incorporates special operations to handle the computing of the least general generalization (lgg) of examples and the code length of logic programs with types. It is possible for previous learners, such as FOIL, GOLEM and Progol, to generate logic programs that include type information represented as is-a relations. However, this expedient has two problems: one in the computation of the code length and the other in the performance. We will illustrate that simply adding is-a relations to background knowledge as ordinary literals causes a problem in computing the code length of logic programs with is-a literals. Experimental results on artificial data show that the learning speed of FOIL exponentially slows as the number of types in the background knowledge increases. The hypotheses generated by GOLEM are about 30% less accurate than those of RHB. Furthermore, Progol is two times slower than RHB. Com- pared to the three learners, RHB can efficiently handle about 3000 is-a relations while still achieving a high accuracy. This in- dicates that type information should be specially handled when learning logic programs with types.