论文部分内容阅读
在现实世界中,障碍物的存在影响了查询点到对象的可见性.可见最近邻查询返回到查询点最近的一个可见对象,是时空数据库中的一类重要应用.由于度量设备的误差和隐私保护,很多关于空间对象位置的数据是不确定的.将不确定对象应用到可见最近邻查询中便产生了概率可见最近邻查询,返回成为可见最近邻概率大于0的对象.有些情况下,用户只关心概率超过一定阈值的结果,于是本文提出了概率阈值可见最近邻查询,返回可见最近邻概率超过阈值τ的不确定对象,其中阈值τ是用户设定的,并且给出了高效的概率阈值可见最近邻查询算法.相比以前的工作,不仅处理了概率和为1的不确定对象,而且处理了概率和小于1的不确定对象;此外,通过引入缺失概率和聚类的概念,提出了高效的过滤技术和快速的批处理技术.最后通过实验验证了本算法的高效性和有效性.
In the real world, the existence of obstructions affects the visibility of the query point to the object, so it can be seen that the nearest neighbor query returns a visible object closest to the query point, which is an important type of application in the spatio-temporal database.Because of the error and privacy Protection, a lot of data on the location of the spatial object is uncertain.Uncertain objects will be applied to the visible nearest-neighbor query will have a probability of visible nearest-neighbor query, return to become the visible nearest neighbor probability greater than 0. In some cases, the user Only concerned with the probability of exceeding a certain threshold of the results, this paper proposes the probability threshold visible nearest neighbor query, returns the visible nearest neighbor probability exceeds the threshold τ of the uncertain object, where the threshold τ is set by the user and gives an efficient probability threshold Compared with the previous work, we not only deal with the uncertain objects with probability 1, but also deal with the uncertain objects with probability less than 1. In addition, by introducing the concept of missing probability and clustering, Efficient filtering technology and fast batch technology.At last, the efficiency and effectiveness of this algorithm are verified through experiments.