优先归纳逻辑程序的极限行为

虽然对归纳逻辑程序的极限行为至今并没有深入的研究，但是通常在分析正在执行的增量式或在线归纳学习算法时，必须考虑这种程序的极限行为．某些归纳学习算法如果不考虑极限行为可能运行到最后会发生错误．如果给定一个递增的例子集合序列，一个归纳逻辑程序会产生一个相应的具有集合论极限的Horn逻辑程序序列，则此归纳逻辑程序是收敛的，并且如果该Horn逻辑程序序列关于例子集合序列的极限是极限正确的，则此归纳逻辑程序是极限正确的，还说明GOLEM系统不是极限正确的．为了解决这个问题，提出了一个极限正确的称为优先GOLEM系统的归纳逻辑系统，并证明了在一定的限制下，优先GOLEM系统的算法是极限正确的．

Limit behavior of prioritized inductive logic programs

Limit behavior of inductive logic programs is an important research topic, but it has not been deeply explored until now. When running incremental or online inductive learning algorithms on a comput er, limit behavior should be taken into account. An example is given to show that some inductive learning algorithms may produce errors in the long run if limit behavior is not considered. An inductive logic pro gram is convergent if, given an increasing sequence of example sets, the inductive logic program creates a corresponding sequence of Horn logic programs which have the set-theoretic limit. Furthermore, it is limit correct if the limit of the sequence of the Horn logic programs is correct with respect to the limit of the sequence of the example sets. Two examples show that the GOLEM system is not limit-correct. Finally, we propose an inductive logic program which is limit-correct, called the prioritized GOLEM system. It is proved that the prioritized GOLEM is limit correct under certain constraints.