面向移动对象连续k近邻查询的双层索引结构

移动对象连续k近邻（CKNN）查询是指给定一个连续移动的对象集合,对于任意一个k近邻查询q,实时计算查询q的k近邻并在查询有效时间内对查询结果进行实时更新.现实生活中,交通出行、社交网络、电子商务等领域许多基于位置的应用服务都涉及移动对象连续k近邻查询这一基础问题.已有研究工作解决连续k近邻查询问题时,大多需要通过多次迭代确定一个包含k近邻的查询范围,而每次迭代需要根据移动对象的位置计算当前查询范围内移动对象的数量,整个迭代过程的计算代价占查询代价的很大部分.为此,提出了一种基于网络索引和混合高斯函数移动对象分布密度的双重索引结构（grid GMM index,GGI）,并设计了移动对象连续k近邻增量查询算法（incremental search for continuous k nearest neighbors,IS-CKNN）.GGI索引结构的底层采用网格索引对海量移动对象进行维护,上层构建混合高斯模型模拟移动对象在二维空间中的分布.对于给定的k近邻查询q,IS-CKNN算法能够基于混合高斯模型直接确定一个包含q的k近邻的查询区域,减少了已有算法求解该区域的多次迭代过程;当移动对象和查询q位置发生变化时,进一步提出一种高效的增量查询策略,能够最大限度地利用已有查询结果减少当前查询的计算量.最后,在滴滴成都网约车数据集以及两个模拟数据集上进行大量实验,充分验证了算法的性能.     

一种针对反向空间偏好top-k查询的高效处理方法

随着地理位置定位技术的蓬勃发展,基于在线位置服务技术的应用也越来越多.提出一种查询类型——反向空间偏好top-k查询.类似于传统的反向空间top-k查询,对于给定的空间查询对象,该查询返回使该对象满足top-k属性得分的那些用户.但不同的是,该对象的属性不是自身具有的特性,而是通过计算该对象与其他偏好对象之间的空间关系（如距离）而确定.这种查询在市场分析等许多重要领域具有需求,例如,根据查询结果,分析出某个地区中某个设施受欢迎的程度.但是,由于大量空间对象的存在导致对象之间空间关系的计算代价非常高,如何实时地计算出对象的空间属性得分,给查询处理带来很大的挑战.针对该问题提出优化的查询处理算法包括：数据集剪枝、数据集批量处理、基于权重的用户分组等策略.通过理论分析和充分的实验验证,证明了所提出方法的有效性.与普通方法相比,这些方法能够大幅度提高查询处理的执行时间和I/O效率.     

一种对时空信息的kNN查询处理方法

互联网上每天都会产生大量的带地理位置标签和时间标签的信息,比如微博、新闻、团购等等,如何在众多的信息中找到在时间和空间地理位置上都满足用户查询需求的信息十分重要.针对这一需求,提出了一种对地理位置和时间信息的k近邻查询（ST-kNN查询）处理方法.首先,利用时空相似度对数据对象的地理位置变量和时间变量进行映射变换,将数据对象映射到新的三维空间中,用三维空间中两点之间的距离相似度来近似代替两个对象之间实际的时空相似度;然后,针对这个三维空间设计了一种ST-Rtree（spatial temporal rtree）索引,该索引综合了空间因素和时间因素,保证在查询时每个对象至多遍历1次;最后,在该索引的基础上提出了一种精确的k近邻查询算法,并通过一次计算确定查询结果范围,从而找到前k个结果,保证了查询的高效性.基于大量数据集的实验,证明了该查询处理方法的高效性.     

可伸缩的增量连续k近邻查询处理

针对基于TPR树(time-parameterized R-tree)索引的大量并发CKNN(continuous k-nearest neighbor)查询处理,提出了一种可伸缩的增量连续k近邻查询处理(scalable processing of incremental continuous k-nearest neighbor queries,简称SI-CNN)框架,通过引入搜索区域进行预裁剪以减少查询更新所需要的TPR树节点访问代价,并引入了增量结果表以保存候选对象,批量地更新查询结果集,具有良好的可伸缩性.基于SI-CNN框架提出了一种增量更新的SI-CNN查询处理算法,能够基于上次查询结果增量的更新查询,支持查询集合中加入或删除查询和移动对象数据集的插入、删除等动态更新操作.实验结果与分析表明,基于SI-CNN框架的SI-CNN算法可以很好地支持大量并发的CKNN查询处理,具有良好的实用价值.     

Top-k查询结果的有界多样化方法

查询结果重复率高是top-k查询处理过程中亟待解决的问题,已有的解决方法需要遍历初始结果集中所有的对象,因此,查询处理的效率较低.为了提高查询处理的效率,把初始结果集映射到欧氏空间中,根据拉式策略,可选用基于得分或基于距离两种方法之一从该空间选出差异最优子空间,在基于距离的方法中,对欧氏子空间进行分割并且利用探测位置和Voronoi图的几何特性减少二次查询对象的数目.在此基础上,提出了top-k查询结果有界多样化算法,并证明了算法的正确性.实验结果表明,所提出的算法提高了top-k查询处理效率.     

基于Voronoi-R*的隐私保护路网k近邻查询方法

针对已有的保护位置隐私路网k近邻查询依赖可信匿名服务器造成的安全隐患,以及服务器端全局路网索引利用效率低的缺陷,提出基于路网局部索引机制的保护位置隐私路网近邻查询方法.查询客户端通过与LBS服务器的一轮通信获取局部路网信息,生成查询位置所在路段满足l-路段多样性的匿名查询序列,并将匿名查询序列提交LBS服务器,从而避免保护位置隐私查询对可信第三方服务器的依赖.在LBS服务器端,提出基于路网基本单元划分的分段式近邻查询处理策略,对频繁查询请求路网基本单元,构建基于路网泰森多边形和R^*树的局部Vor-R^*索引结构,实现基于索引的快速查找.对非频繁请求路网基本单元,采用常规路网扩张查询处理.有效降低索引存储规模和基于全局索引进行无差异近邻查询的访问代价,在保证查询结果正确的同时,提高了LBS服务器端k近邻查询处理效率.理论分析和实验结果表明,所提方法在兼顾查询准确性的同时,有效地提高了查询处理效率.     

基于图压缩的k可达查询处理

研究了基于图压缩的k可达查询处理,提出了一种支持k可达查询的图压缩算法k-RPC及无需解压缩的查询处理算法,k-RPC算法在所有基于等价类的支持k-reach查询的图压缩算法中是最优的.由于k-RPC算法是基于严格的等价关系,因此进一步又提出了线性时间的近似图压缩算法k-GRPC.k-GRPC算法允许从原始图中删除部分边,然后使用k-RPC获得更好的压缩比.提出了线性时间的无需解压缩的查询处理算法.真实数据上的实验结果表明,对于稀疏的原始图,两种压缩算法的压缩比分别可以达到45%,对于稠密的原始图,两种压缩算法的压缩比分别可以达到75%和67%;与在原始图上直接进行查询处理相比,两种基于压缩图的查询处理算法效率更好,在稀疏图上的查询效率可以提高2.5倍.     

基于邻域k-核的社区模型与查询算法

现实生活中的网络通常存在社区结构,社区查询是图数据挖掘的基本任务.现有研究工作提出了多种模型来识别网络中的社区,如基于k-核的模型和基于k-truss的模型.然而,这些模型通常只限制社区内节点或边的邻居数量,忽略了邻居之间的关系,即节点的邻域结构,从而导致社区内节点的局部稠密性较低.针对这一问题,本文将节点的邻域结构信息融入k-核稠密子图中,提出一种新的基于邻域连通k-核的社区模型,并定义了社区的稠密度.基于这一新模型,研究了最稠密单社区搜索问题,即返回包含查询节点集且具有最高稠密度的社区.在现实生活图数据中,一组查询节点可能会分布在多个不相交的社区中.为此,本文进一步研究了基于稠密度阈值的多社区搜索问题,即返回包含查询节点集的多个社区,且每个社区的稠密度不低于用户指定的阈值.针对最稠密单社区搜索和基于稠密度阈值的多社区搜索问题,首先定义了边稠密度的概念,并提出了基于边稠密度的基线算法.为了提高搜索效率,设计了索引树和改进索引树结构,能够支持在多项式时间内返回查询结果.通过与基线算法在多组数据集上的对比,验证了基于邻域连通k-核的社区模型的有效性和所提出查询算法的效率.     

一种支持Top-k空间关键词检索的高效压缩索引

基于位置的服务可以指引用户找到在特定位置或区域内能够提供所需要服务的对象(比如找某个高校附近(经纬度标识)的咖啡店).向这类服务提交一个查询位置和多个关键词,该类服务返回k个最相关的对象,对象和查询的相关性同时考虑空间相近性和文本相似性.为了支持高效的top-k空间关键词查询,出现了多种混合索引,然而现有的这些索引为了提供实时响应均耗费大量存储空间.提出一种基于压缩技术的索引CSTI,该索引显著减少了存储开销(至少减少80%甚至到两个数据量级),同时保持高效的查询性能.大量基于真实和仿真数据集的实验结果表明,CSTI在空间开销和响应时间上均优于已有方法.     

带间隔约束的Top-k对比序列模式挖掘

对比序列模式能够表达序列数据集合间的差异,在商品推荐、用户行为分析和电力供应预测等领域有广泛的应用.已有的对比序列模式挖掘算法需要用户设定正例支持度阈值和负例支持度阈值.在不具备足够先验知识的情况下,用户难以设定恰当的支持度阈值,从而可能错失一些对比显著的模式.为此,提出了带间隔约束的top-k对比序列模式挖掘算法kDSP-Miner(top-k distinguishing sequential patterns with gap constraint miner).kDSP-Miner中用户只需设置期望发现的对比最显著的模式个数,从而避免了直接设置对比支持度阈值.相应地,挖掘算法更容易使用,并且结果更易于解释.同时,为了提高算法执行效率,设计了若干剪枝策略和启发策略.进一步设计了kDSP-Miner的多线程版本,以提高其对高维序列元素情况的处理能力.通过在真实世界数据集上的详实实验,验证了算法的有效性和执行效率.     

Direction-aware KNN queries for moving objects in a road network

Recently more and more people focus on k-nearest neighbor (KNN) query processing over moving objects in road networks, e.g., taxi hailing and ride sharing. However, as far as we know, the existing k-nearest neighbor (KNN) queries take distance as the major criteria for nearest neighbor objects, even without taking direction into consideration. The main issue with existing methods is that moving objects change their locations and directions frequently over time, so the information updates cannot be processed in time and they run the risk of retrieving the incorrect KNN results. They may fail to meet users’ needs in certain scenarios, especially in the case of querying k-nearest neighbors for moving objects in a road network. In order to find the top k-nearest objects moving toward a query point, this paper presents a novel algorithm for direction-aware KNN (DAKNN) queries for moving objects in a road network. In this method, R-tree and simple grid are firstly used as the underlying index structure, where the R-tree is used for indexing the static road network and the simple grid is used for indexing the moving objects. Then, it introduces the notion of “azimuth” to represent the moving direction of objects in a road network, and presents a novel local network expansion method to quickly judge the direction of the moving objects. By considering whether a moving object is moving farther away from or getting closer to a query point, the object that is definitely not in the KNN result set is effectively excluded. Thus, we can reduce the communication cost, meanwhile simplify the computation of moving direction between moving objects and query point. Comprehensive experiments are conducted and the results show that our algorithm can achieve real-time and efficient queries in retrieving objects moving toward query point in a road network.

     

k-Nearest Neighbor Query Processing Algorithms for a Query Region in Road Networks

Recent development of wireless communication technologies and the popularity of smart phones are making location-based services (LBS) popular. However, requesting queries to LBS servers with users’ exact locations may threat the privacy of users. Therefore, there have been many researches on generating a cloaked query region for user privacy protection. Consequently, an effcient query processing algorithm for a query region is required. So, in this paper, we propose k-nearest neighbor query (k-NN) processing algorithms for a query region in road networks. To effciently retrieve k-NN points of interest (POIs), we make use of the Island index. We also propose a method that generates an adaptive Island index to improve the query processing performance and storage usage. Finally, we show by our performance analysis that our k-NN query processing algorithms outperform the existing k-Range Nearest Neighbor (kRNN) algorithm in terms of network expansion cost and query processing time.     

基于Voronoi图的连续反向最近邻查询

为解决动态环境中移动点的连续反向最近邻查询问题,将连续反向最近邻查询分为单色和双色2种情况进行研究。利用移动点Voronoi图,分别给出单色连续反向最近邻查询算法、双色连续反向最近邻查询算法以及相关定理,对算法正确性和可终止性进行证明,分析算法时间复杂性。按照移动点Voronoi图的拓扑结构是否改变分为2种情况,分析每种情况下候选所在区域的变化,在变化区域内进行Voronoi图的重构,得到对应的解决方法。在多数情况下,该算法只需生成局部移动点的Voronoi图即可找到结果,减小了连续反向最近邻查询的代价。     

Nearest and reverse nearest neighbor queries for moving objects

With the continued proliferation of wireless communications and advances in positioning technologies, algorithms for efficiently answering queries about large populations of moving objects are gaining interest. This paper proposes algorithms for k nearest and reverse k nearest neighbor queries on the current and anticipated future positions of points moving continuously in the plane. The former type of query returns k objects nearest to a query object for each time point during a time interval, while the latter returns the objects that have a specified query object as one of their k closest neighbors, again for each time point during a time interval. In addition, algorithms for so-called persistent and continuous variants of these queries are provided. The algorithms are based on the indexing of object positions represented as linear functions of time. The results of empirical performance experiments are reported.     

An Enhanced Technique for k-Nearest Neighbor Queries with Non-Spatial Selection Predicates

In multimedia databases, k-nearest neighbor queries are popular and frequently contain non-spatial predicates. Among the available techniques for such queries, the incremental nearest neighbor algorithm proposed by Hjaltason and Samet is known as the most useful algorithm [16]. The reason is that if k > k neighbors are needed, it can provide the next neighbor for the upper operator without restarting the query from scratch. However, the R-tree in their algorithm has no facility capable of partially pruning tuple candidates that will turn out not to satisfy the remaining predicates, leading their algorithm to inefficiency. In this paper, we propose an RS-tree-based incremental nearest neighbor algorithm complementary to their algorithm. The RS-tree used in our algorithm is a hybrid of the R-tree and the S-tree, as its buddy tree, based on the hierarchical signature file. Experimental results show that our RS-tree enhances the performance of Hjaltason and Samet's algorithm.     

Continuous reverse k nearest neighbors queries in Euclidean space and in spatial networks

In this paper, we study the problem of continuous monitoring of reverse k nearest neighbors queries in Euclidean space as well as in spatial networks. Existing techniques are sensitive toward objects and queries movement. For example, the results of a query are to be recomputed whenever the query changes its location. We present a framework for continuous reverse k nearest neighbor (RkNN) queries by assigning each object and query with a safe region such that the expensive recomputation is not required as long as the query and objects remain in their respective safe regions. This significantly improves the computation cost. As a byproduct, our framework also reduces the communication cost in client–server architectures because an object does not report its location to the server unless it leaves its safe region or the server sends a location update request. We also conduct a rigid cost analysis for our Euclidean space RkNN algorithm. We show that our techniques can also be applied to answer bichromatic RkNN queries in Euclidean space as well as in spatial networks. Furthermore, we show that our techniques can be extended for the spatial networks that are represented by directed graphs. The extensive experiments demonstrate that our techniques outperform the existing techniques by an order of magnitude in terms of computation cost and communication cost.     

Visible Reverse k-Nearest Neighbor Query Processing in Spatial Databases

Reverse nearest neighbor (RNN) queries have a broad application base such as decision support, profile-based marketing, resource allocation, etc. Previous work on RNN search does not take obstacles into consideration. In the real world, however, there are many physical obstacles (e.g., buildings) and their presence may affect the visibility between objects. In this paper, we introduce a novel variant of RNN queries, namely, visible reverse nearest neighbor (VRNN) search, which considers the impact of obstacles on the visibility of objects. Given a data set P, an obstacle set O, and a query point q in a 2D space, a VRNN query retrieves the points in P that have q as their visible nearest neighbor. We propose an efficient algorithm for VRNN query processing, assuming that P and O are indexed by R-trees. Our techniques do not require any preprocessing and employ half-plane property and visibility check to prune the search space. In addition, we extend our solution to several variations of VRNN queries, including: 1) visible reverse k-nearest neighbor (VRkNN) search, which finds the points in P that have q as one of their k visible nearest neighbors; 2) delta-VRkNN search, which handles VRkNN retrieval with the maximum visible distance delta constraint; and 3) constrained VRkNN (CVRkNN) search, which tackles the VRkNN query with region constraint. Extensive experiments on both real and synthetic data sets have been conducted to demonstrate the efficiency and effectiveness of our proposed algorithms under various experimental settings.