Metarule Replacement Datasets

该数据集用于进行元规则替换实验，展示了自动推导的排序元规则可以替代用户定义的排序元规则，而且不会影响预测准确性和训练时间。此外，该数据集与涉及不同元规则替换阶段的实验相关联，并用于评估TOIL系统的性能表现。这项任务主要针对元解释学习中的元规则替换。

This dataset is developed for meta-rule replacement experiments, demonstrating that automatically derived ranking meta-rules can replace user-defined ranking meta-rules without compromising prediction accuracy or training time. Furthermore, this dataset is linked to experiments involving different stages of meta-rule replacement, and is utilized to evaluate the performance of the TOIL system. This task primarily focuses on meta-rule replacement in meta-explanation learning.