为了实时监控路网上移动对象(车辆)的运动,各移动对象不断向中心服务器汇报其位置,中心服务器存储数据以响应用户的各种查询。此类方法不仅通信开销巨大,增加服务器负载,而且不能同时满足群体态势感知和个体移动对象位置追踪的需求。因此,提出一种基于时空锚点的双粒度移动感知(Double-granularity Movement Detection Based on Spatial-temporal Anchors,DMDSA)框架,将移动对象嵌入时空网格,其经过时空锚点时向服务器汇报其运动模式,实现对群体运动的感知和个体移动的追踪。离线阶段,服务器从历史轨迹中挖掘运动模式;移动对象运动时,服务器结合挖掘的运动模式,在线计算聚合模式表征群体运动,并采用最大似然估计确定目标的运动模式,实现群体态势感知;进一步,采用锚点独立策略和锚点序列策略识别最可能的运动序列,实时追踪个体对象的运动。在模拟数据集和实际数据集上的实验表明,所提方法在大幅度减小位置汇报代价的前提下,不仅能够准确地监控区域的群体运动态势,并且能够有效地追踪和预测个体移动对象的位置,有助于智慧城市的建设。
Website navigability is acquiring a growing importance in website design and redesign,quality evaluation,and improvement.Existing navigability measures mainly depend on site link structure,so that they only consider the impact of site link structure for navigability and ignore the impact of Web page content.A continuous Markov chain model which depicts the user's surfing behavior can balance these two factors in the evaluation of website navigability,and it needs to estimate the page transition probabilities and user stay time according to user access log.In this way,we can obtain more reliable results for website navigability measure than the existed methods.Experiments show that our method is effective.
