The positional accuracy of crowdsourced geographic data from open source web GIS

Abstract

Geographic information and data are becoming more important in today modern society. One popular and widely used method to obtain geographic data is by crowdsourcing. The positional accuracy of geographic data obtained in this way; specifically, one obtained without the help of positioning devices are evaluated in this paper. The crowdsourced data used in this paper were obtained from a web mapping application built with open source tools. This allows many parameters to be collected during the crowdsourcing process that helps finding factors related to positional accuracy. It is found that each crowdsourced place has a different tendency to be located accurately on a map. Factors like residence time, screen sizes, and feature distance to street junctions were found to be related with positional accuracy. These findings could be used to make better decision in the crowdsourcing process.