兴趣点数据的组织模型与检索算法

The morphological model and retrieving algorithm of the POI data

兴趣点(POI)数据是现代电子导航地图的主要组成部分,有效的数据组织模型在满足快速检索的前提下应尽可能地节省存储空间,且高效检索算法是导航电子地图实时应用中的关键技术。该文针对兴趣点的位置和属性信息产生唯一的ID编码,并进行功能分区,将分区信息表与兴趣点存储偏移量直接关联,通过机械分词和多叉树,建立字典树,实现兴趣点数据的有序组织,通过分层偏移、逐字映射实现快速检索。采用该方法对约86万条名称数据进行测试,结果表明,存储结构好,响应速度快。

Point of interest（POI）data is the key contents of modern digital navigation maps,agood organization model should save space and be easy to use,and the efficient retrieving algorithm should fulfill the requirement of the real-time navigation,which play important roles in the application of POI data.In this paper,a unique ID code for an object was generated from the information of the POI＇s location and other attributes according to the Hash table rule,the POI data were partitioned into four classes,which are name data,location data,number data,and category data.The links between the partition information table and POI storage offset were established.POI data was orderly organized with the lexicographic tree which was created through the multi branches tree algorithm and the mechanical segmentation of POI data,and fast retrieving was realized through hierarchical migration and word-by-word mapping.A test was done which includes about 860 000 POI name data,the test＇s results showed that the proposed method had a good performance in POI data organization and retrieval.