中文摘要：

为了及时准确地获取用于智慧位置服务的城市层次性空间知识,该文提出了一种依据签到属性显著度的差异从位置签到数据中提取城市分层地标的方法.首先,针对传统基于调查问卷的公众认知度计算方法存在的不足,该文从位置签到数据中的签到次数、签到用户数和用户影响因子3个方面入手讨论POI显著度的影响因素,进而提出了一种基于签到数据的POI显著度计算模型;然后,利用Voronoi图对POI空间作用范围进行分割并结合邻域分析与合并算法逐层迭代,形成了基于签到POI显著度与Voronoi图的分层地标提取算法;最后,以北京市的位置签到数据为例,进行了POI显著度计算与分时分层地标提取,通过与现有实体、网络地图资料的比对,验证了算法的有效性,并进一步分析了分层、分时地标反映的空间现象和规律,为个性化智慧位置服务提供城市层次性空间知识.

英文摘要：

For acquiring the city hierarchical spatial knowledge to be applied in smart location service on time, a method of extracting city hierarchical landmarks from check-in data according to their check-in attribute significance is proposed. To overcome the disadvantage of traditional hierarchical landmarks extracting method of public cognition based on questionnaires, a POI significance measure model is constructed after analyzing the factors influencing the significances of POI objects from three vectors which are check-in numbers, check-in users, user impact factor in check-in data. Then, combined with neighborhood analysis and merge algorithms, the POI space scope is segmented into Thiessen polygons with Voronoi diagrams iteratively, that is the hierarchical landmark extracting algorithm based on POI significance and Voronoi diagrams. An experiment is carried out to compute the significances of the POIs selected from the area of Beijing City, and the POI in the Thiessen polygon with higher significance than its surroundings are treated as landmarks in each level at last. The paper confirms the efficiency of the algorithm by comparing the hierarchical landmarks with real objects and electronic map data, analyzes the spatial phenomenon and regularity reflected by the hierarchical time-sharing landmarks got from check-in data. This can be used to provide hierarchical spatial knowledge of city for personalized intelligent location service.