移动对象频繁周期模式发现

Mining Periodic-frequent Patterns of Moving Objects

随着信息技术的快速发展,挖掘移动对象背后隐藏的模式越来越重要.为了解决现有方法无法区分移动对象数据中的频繁项和稀有项以及不能满足向下闭包、挖掘效率低等问题,提出一种具有向下闭包特性的闭包多限制条件树算法(MultiConstraint Closure Conditional Tree,MCCCT).算法为不同模式设置不同限制条件,解决了组合爆炸和稀有项问题,提高了挖掘效率.针对模式支持度和周期距离难于获取的问题,根据每个模式出现频数动态获取属性值,增强了挖掘算法的灵活性.为防止噪音等不确定因素的影响,引入基于相似度的模式匹配算法,使模式挖掘更加具有健壮性.采用公开移动对象数据进行实验,结果表明算法能高效挖掘出的频繁周期模式.

With the rapid development of the information technology, It is more import to mine patterns behind these moving data. The existing methods cannot distinguish between the frequent items and rare items and it is inefficient. So, we present an algorithm which called Multi-Constraint Closure Conditional Tree （ MCCCT）. The algorithm set different limiting conditions to solve the com- binatorial explosion problem and the rare item problem and it is more efficient. Because people cannot know each pattern＇s support and period, we obtain patterns＇ supports and periods dynamically which enhanced algorithm flexibility. To prevent noise and other un- certainties, we introduced similarity-based pattern matching method, making the method more robust. The experimental results with open data show that this method can be efficiently excavated the Deriodic-freauent oatterns.