GFExtractor：事件序列上有效挖掘无冗余情节规则的算法

摘要

事件序列上挖掘情节规则，旨在发现情节之间的因果关系。基于非重叠的最小发生的支持度定义及深度优先搜索策略，提出在事件序列上挖掘无冗余情节规则的GFExtractor算法。利用非生成子情节的剪枝策略，淘汰非生成子情节；利用向前、向后扩展检查，淘汰非闭情节；最终在情节生成子集Gen与频繁闭情节集FCE之间产生无冗余的情节规则。实验结果证实了算法在事件序列上挖掘无冗余情节规则的有效性。%Mining episode rules in event sequence aims to discover the causal relationship between the episodes. To mine non-redundant episode rules in event sequence, the algorithm of GFExtractor is proposed in this paper, based on the support defi-nition of non-overlapping minimal occurrences and the depth-first search strategy. GFExtractor uses the pruning technology to eliminate non-generator episodes, and uses the forward and backward extension check to eliminate non-closed episodes. Non-redundant episode rules are generated between a superset of Gen and FCE. Experimental results confirm the validity of algo-rithm in mining non-redundant episode rules in event sequence.