Covering vs. Divide-and-Conquer for Top-Down Induction of Logic Programs

Covering and divide-and-conquer are two wellestablished search techniques for top-down in­ duction of propositional theories However, for top-down induction of logic programs, only covering has been formalized and used extensively In this work, the divide-and-conquer technique is formalized as well and compared to the covering technique in a logic program­ ming framework Covering works by repeat­ edly specializing an overly general hypothe­ sis, on each iteration focusing on finding a clause wi th a high coverage of positive exam­ ples Divide-and-conquer works by specializ­ ing an overly general hypothesis once, focus­ ing on discriminating positive from negative examples Experimental results are presented demonstrating that there are cases when more accurate hypotheses can be found by divideand-conquer than by covering Moreover, since covering considers the same alternatives repeat­ edly it tends to be less efficient than divideand-conquer, which never considers the same alternative twice On the other hand, cover­ ing searches a larger hypothesis space, which may result in that more compact hypotheses are found by this technique than by dmde-andconquer Furthermore, divide-and-conquer is, in contrast to covering, not applicable to learn­ ing recursive definitions,