Measuring the Spatial Similarities in Volunteered Geographic Information

Abstract

Abstract. Volunteered citizens have the potential to be used as social and distributed sensors, monitoring their surroundings and producing and sharing massive amounts of geographic data. The degree of spatial similarities in Volunteered Geographic Information (VGI) somehow indicates an index for the accuracy and precision of user-generated spatial data. In other words, the spatial similarities in VGI refer to how close a citizen-generated spatial feature is to the true (or accepted one) or how close the citizen-generated spatial features of the same geographic phenomenon are to each other. The present study aims at developing a Web-based GIS tool to collect VGI and extract the spatial similarity indexes. To this end, a case study involving the identification of optimal areas for restaurants in Babolsar, Mazandaran province was used. The degree of similarity between the areas (polygons) proposed by citizens was then investigated using spatial indicators of intersection and minimum central distance. The results show that with the increase in the frequency of citizen-generated polygons, the geometric dispersion of the polygons decreases, and the similarity of citizens’ polygons to establish a restaurant increases. With the increasing agreement, the amount of standard deviation in the area, perimeter, and minimum central distance of intersection areas reduces from 4645 to 15.4, 134.5 to 21.6, and 42.4 to 4.2, respectively.

An Aggregated Shape Similarity Index: A Case Study of Comparing the Footprints of OpenStreetMap and INSPIRE Buildings

“…However, the evaluation of the positional accuracy of areal objects is much more complex. For example, in the articles [28][29][30] the distance between the centroid of two polygons was used as a measure of similarity or its simple part. However, this parameter does not uniquely identify objects and does not sufficiently quantify their similarity.…”