干预规则挖掘是近年从干预实践中提出的新型数据挖掘任务,旨在利用数据挖掘技术探测干预事件,发现最佳干预时机和力度,提供促进事物向期待状态转化的决策支持.文中以四年的研究实践为背景,介绍干预规则挖掘的研究沿革和现状,给出了干预规则挖掘的任务分类.从三个角度,即干预效果预测、干预方法发现和未知干预探测三方面,介绍干预规则挖掘的研究问题、困难和成果.展望了干预规则挖掘未来研究方向.%Intervention rule mining is an emerging data mining task, which is derived from the practice of intervention application. It aims at applying data mining techniques on detecting intervention events, discovering the best intervention time and intensity, and decision support for converting objects from undesirable state to desirable state. This paper introduces the research background, as well as the major related advances on intervention rule mining based on the four-year practice, and defines the task classification. Moreover, this paper surveys the research issues, difficulties and achievements in three aspects, i. e. intervention effect prediction, intervention method discovery, and unknown intervention event detection. Finally, this paper discusses the future work of intervention rule mining.
展开▼
