基于时序的不同事物同属性的关联规则挖掘

The Association Rule Mining of Same Attributes but Different Things Based on the Temporal Logic

基于时序的关联规则挖掘算法的研究一直都是人们关注的课题,提出了一种基于时序逻辑的不同事物同属性的关联规则挖掘。传统的关联规则主要是揭示了多个事物的同一属性在相同的时间点上的相互关联性,这样的关联规则的项与项之间没有体现时间上的差别,也就无法对时间序列的发展趋势进行预测。实验表明这种方法对于不同事物同属性预测具有现实意义。

The timing-based association rule mining algorithm has always been the subject of attention, this paper presents a temporal logic based on the different things the same attribute association rules mining. Traditional association rules is to reveal the interconnectedness of things the same property on the same point in time, this association rules between items does not reflect the time difference, it can not be on the development of time series The trend forecast.Association rules mining association rule mining method based on temporal logic of different things with the property reflects the degree of corre- lation between the different states of the same attributes of different things on the different points in time. The result shows this algorithm is of practical significance on single-property prediction things.