基于POI数据的模糊地名空间范围估计

在信息检索中常用地名作为空间参照,因此确定模糊地名的空间范围备受关注。由于模糊地名不存在确切的边界,评估模糊地名空间范围十分困难,但人们对模糊地名的认知区域可以提供参考。本文分析了模糊地名空间的基本特征;以案例形式,讨论了利用POI数据和认知实验确定模糊地名空间范围的基本方法和特点,并分析这些案例地名空间的特性。从这些案例研究可以得出以下结论:影响模糊地名空间范围的因素有道路、熟悉程度和地名借用;用POI数据对模糊地名估计出的空间范围符合人们认知的空间范围。

Modeling vague places using point of interest(POI)datasets

Places are often used as references in geographic information retrieval,therefore,defining approximate regions for vague places attracts much interest.A well-known problematic issue is to evaluate modeled regions for vague places as no authority exists that defines such regions.Cognitive regions for vague places can be used as reasonable data source for evaluation.In this paper,main features of vague places are summarized,in which the regions of 5 vague places in Wuhan city are modeled using POI datasets and demarcated by questionnaire.The spatial characteristics of the vague places are discussed.It is found that regions as the study cases are influenced by roads,people's familiarity and name borrowing,and the modeled results using POI datasets are in line with people's cognition.