Extracting hierarchical landmarks from urban POI data

Abstract

For acquiring the hierarchical spatial knowledge to be applied in cognitive route directions, a method of extracting hierarchical landmarks from urban POI data according to their signifi cances is proposed. After analyzing the factors infl uencing the signifi cances of POI objects from public cognition, spatial distribution and individual characteristics, a signifi cance measure model composed of three vectors which are public cognition degree, urban centrality degree and characteristic attribute value is constructed. Then, the processes of computing the vector values of POI objects are discussed by the methods of questionnaire survey, multi-density spatial clustering and data normalization respectively. An experiment is carried out to compute the signifi-cances of the POIs selected from the area of Wuchang region of Wuhan city, and the POIs with different signifi cances are treated as landmarks in different levels at last. In this experiment, several levels of landmarks are extracted, and being used as seeds to compute weighted Voronoi diagrams in every level, to refl ect the infl uence area of every landmark and associate the landmarks in the same level and between the sequential levels.