分布式空间数据库中有效数据频繁项实时检测

Real Time Detection of Active Data Frequent Items in Distributed Spatial Database

当前数据频繁项检测方法往往具有数据存储结构复杂、维护困难和复杂度高的弊端;针对这种情况,提出一种新的分布式空间数据库中有效数据频繁项实时检测方法。对事物数据进行分配,建立全局VFP树,将最近窗口中全部项集添加至全局VFP树,通过临时表对项集信息进行保存,利用遍历全局树实时检测有效数据的频繁项。通过滑动窗口的初始化阶段与滑动阶段对最旧一批项集临时表进行处理,实现维护。实验结果表明,所提方法能够有效实现分布式空间数据库中有效数据频繁项的实时检测;和其他方法相比,复杂度低、检测可靠性高。

Current data frequent item detection methods often have the disadvantages of complex data storage structure,difficult maintenance and high complexity.Aiming at this situation,a new real-time detection method for effective data frequent items in distributed spatial database was proposed.Data was assigned to the objects,the global VFP tree was established,all the itemsets in the nearest window was added to the global VFP tree,the itemsets information through the temporary table was preserved,and the frequent items of effective data in real time was detected by traversing the global tree.The maintenance is realized by processing the temporary table of the oldest set of items by the initialization and sliding stages of the sliding window.The experimental results show that the proposed method can effectively detect the frequent items in distributed spatial database.Compared with other methods,it has low complexity and high detection reliability.