路网中空间关键字（连续）k近邻查询算法研究

项目名称： 路网中空间关键字（连续）k近邻查询算法研究

项目编号： No.61309002

项目类型： 青年科学基金项目

立项/批准年度： 2014

项目学科： 自动化技术、计算机技术

项目作者： 李艳红

作者单位： 中南民族大学

项目金额： 23万元

中文摘要： 位置相关查询处理(LBQs)作为位置相关服务(LBSs)的支持性技术之一，已经成为一个研究热点问题。然而现有的位置相关查询算法多数仅仅考虑了查询点和待查询对象的位置信息，没有考虑待查询对象与查询的关键字相似程度，不能完全满足实际应用的需要。近年来，研究者提出了将这两种因素相结合的新型查询处理类型，即空间关键字查询。目前，已有的空间关键字查询研究成果大多局限于欧氏空间，属于对现实情况的简化。由于欧氏空间的相关算法不能通过简单的修改而应用到路网中，项目研究路网中空间关键字查询处理策略。综合考虑路网距离与关键字相似度这两个因素，对路网结构、路网上的对象进行合理的划分、组织，并提出适当的索引结构。提出灵活的相似度评价模型，使用户能够根据实际需要来设置路网距离、关键字相似程度这两个因素的权重。设计实现路网中空间关键字(连续)k近邻查询算法。优化所设计的算法，提高系统的可扩展性、查询结果时效性。

中文关键词： 时空数据库；查询处理；算法；空间关键字查询；路网

英文摘要： As one of the enabling technologies for location-based services(LBSs), location-based queries (LBQs) have become a hot research topic. However, most of the existing LBQ methods only consider the location information of the query point and data objects, while ignoring the keyword similarity between them. Thus, these methods can not completely satisfy actual applications. Recently, the spatial keyword query, which is a combination of a keyword query and a spatial query, has been proposed by researchers. However, existing research works on spatial keyword query processing are almost limited in Euclidean space, which is a simplification of realistic scenarios. Since the query methods in Euclidean space can not be easily extended to handle queries in road networks, we mainly address the problem of processing spatial keyword queries in road networks. Firstly, by taking into account both the location information and the keyword information of the query point and objects, the road network structure and the data objects in it will be partitioned and organized efficiently, and an efficent road network index structure will be proposed. Secondly, a flexible similarity evaluation model will be presented, thus the users can freely choose the weights of network distance and keyword similarity, respectively. Thirdly, efficient

英文关键词： Spatio-Temporal Database；Query process；Algorithm；Spatial keyword queries；Road networks