再论关系数据流中增减型及循环型知识的发现

Discovering knowledge of increase or decrease and knowledge of relational data stream

继续研究在以关系为单位的数据流中发现"增加"、"减少"等增减型知识以及"良性循环"、"恶性循环"等循环型知识的方法。这些方法可以发现事物间发展的联系及发展的趋势,为决策者提供更强有力的支持。特别是发现各种循环,将为决策者加强"良性循环",增加效益与成功,摆脱"恶性循环",避免亏损与失败,提供了重要的依据。

This paper continues the research of method for discovering knowledge with ＂increase＂ or ＂decrease＂ properties and ＂circle＂ properties （benign circle or vicious circle） in data stream that its unit is relation. Discovering algorithm of the two kinds of decision rule in increase-decrease type or cycle type can be used to discover the developmental connection between objects and the trend of development. And then we present the similarity of rules used to decide the providing occasion of rule for decision maker. Specially, discovering all kinds of circle can provide important foundation for decision maker to enforce benign circle, increase benefit and success, and get rid of vicious circle to avoid loss and failure.