Rank-based Metric Adjustments

This table contains the adjustments for rank-based metrics for datasets in PyKEEN stratified by split (i.e., training, testing, validation) as well as ranking side (i.e., head, tail, both). If you're not familiar with ranking sides, you probably want <em>both</em>. The columns in the TSV are: dataset metric side split value本表收录了PyKEEN中各数据集基于排序的评价指标的校正结果，校正结果按数据集划分（即训练集、测试集、验证集）与排序侧（即头实体、尾实体、双侧）进行分层。若您不熟悉排序侧的相关概念，建议选择「双侧」。制表符分隔值（Tab-Separated Values，TSV）文件的列依次为：数据集、指标、排序侧、划分方式、指标值。