基于周期规律的加权时态关联规则挖掘算法

数据挖掘过程中只考虑数据项权重或者只考虑时态语义会导致挖掘结果不全面。针对该问题，对加权关联规则、时态关联规则和时态数据周期规律进行研究，将权值、K-支持期望和周期等概念引入到时态关联规则中，提出一种基于周期规律的加权时态关联规则挖掘算法。以某管理系统审计数据为例进行实验验证，结果表明该算法能够准确地挖掘出数据库中的加权时态关联规则，与加权关联规则算法相比，在时间复杂度相同的情况下能使关联规则的挖掘结果更加全面。

Weighted Temporal Association Rule Mining Algorithm Based on Periodic Law

In data mining process, it only considers the weight of data items, or only considers the temporal semantics of data, resulting in incomplete mining results. For the above-mentioned problem, the weighted association rules, temporal association rules and the periodicity of temporal data are studied. The weights, K-support bounds and periodicity are introduced into temporal association rules. The algorithm of mining weighted temporal association rules based on periodicity is proposed. Applying the algorithm in audit data of a management system, results show that this algorithm can accurately mine the weighted temporal association rules, and with the same time complexity, it makes the mining of association rules more meticulousness than weighted association rule algorithm.