一种面向集合的空间关键词查询方法研究

发布时间：2018-11-26 09:25

【摘要】：如今,定位技术的快速发展使得基于位置的服务得到了广大移动用户的青睐。随之而来的,各种拥有定位技术的移动设备不断推陈出新,各种基于位置信息和文本信息的查询被提出。为了满足这样的查询需求,学术界对于不同条件的查询需求分别提出了各种高效的空间关键词查询算法。空间关键词查询按类别分为,单一对象匹配的空间关键词查询,面向集合的空间关键词查询,面向集合和方向的空间关键词查询。不同类型的查询,在不同的查询条件或不同的查询环境中,也有各种不同的查询算法。道路网是人们生活空间的抽象,研究道路网上的空间关键词查询更有实际意义和使用价值。到目前为止,人们还没有研究过道路网上方向和集合同时被考虑的空间关键词查询。在前人的研究基础上,本文提出了一种基于道路网的面向集合和方向的空间关键词查询方法。首先,把整个道路网分割成几个相连的局域网,以局域网层次的结构建立空间索引,为每个局域网建立倒排索引,目的是每次查询时,可以快速的跳过不包含查询关键词的区域直接进入下一个包含查询关键词的局域网进行查询。然后,找到在指定查询方向上并且在指定查询范围内的局域网。最后,在获得可查询的局域网内,查询出满足查询文本信息和空间位置信息及查询范围约束的一条路线,这条路线满足:1)路线上的空间对象文本信息能够共同满足查询的文本信息;2)此路线是最优的路线,即查询用户从查询位置出发,走此路线能把所要做的事情完成,并且是代价(可以是行走的路程或者行走的时间或者所消耗的费用等)最小。本文针对道路网上面向集合且受查询方向约束的空间关键词查询,提出了面向无主关键词的查询和主关键词优先的查询。无主关键词的查询即从查询点出发按照道路网在可查询范围内扩展查询。主关键词优先的查询,首先在可查询范围内以一种迭代替换的方式进行扩展查询直到查询到主关键词对象,若还有关键词没有被已查询到的空间对象所覆盖,则以面向无主关键词的查询方式继续进行扩展查询。本文对这两种查询分别给出了查询算法。最后,分别对以上两种查询进行了实验,证明了所提算法的有效性。
[Abstract]:Nowadays, with the rapid development of location technology, location-based services are favored by mobile users. As a result, various mobile devices with location technology are constantly emerging, and various queries based on location information and text information are proposed. In order to meet this query demand, the academic community proposed a variety of efficient spatial keyword query algorithms for different query requirements. Spatial keyword query is divided into categories, single object matching spatial keyword query, set and direction oriented spatial keyword query. There are different query algorithms in different query conditions and different query environments. Road network is the abstraction of people's living space. Up to now, people have not studied the spatial keyword query that the direction and set of road network are considered simultaneously. On the basis of previous researches, this paper proposes a method of spatial keyword query based on road network oriented to set and direction. First of all, divide the whole road network into several connected local area networks, build spatial index with the structure of local area network, build inverted index for each local area network, aim at every query, Can quickly skip the region without query keywords directly into the next local area network containing query keywords for query. Then, find the local area network in the specified query direction and within the specified query scope. Finally, in the local area network that can be queried, a route satisfying the query text information, spatial location information and query range constraints is found. This route satisfies: 1) the spatial object text information on the route can satisfy the query text information together; 2) this route is the best route, that is to say, the query user starts from the query position, and takes this route to complete what needs to be done, and it is the least cost (it can be the distance or time spent or the expense consumed). Aiming at the spatial keyword query which is set oriented and constrained by the query direction on the road network, this paper proposes a query oriented to the main keyword and a query with the primary keyword first. The query without main key words starts from the query point and expands the query according to the road network within the range of query. In the main keyword first query, the query is extended by an iterative replacement method within the query scope until the main keyword object is queried, if the keyword is not overwritten by the queried spatial object. Then the extended query is continued in a way that is oriented to the key words. In this paper, the two query algorithms are given respectively. Finally, experiments on the above two kinds of queries are carried out to prove the effectiveness of the proposed algorithm.
【学位授予单位】：哈尔滨工程大学
【学位级别】：硕士
【学位授予年份】：2016
【分类号】：TP301.6

【参考文献】