Handling quality in crowdsourced geographic information

Abstract

While spatial information quality is an established discipline in traditional scientific geographical information (GI), standards and protocols for representing and assessing the quality of geographic contributions generated by volunteers or by the generic ‘web crowd’ are still missing. This work offers an analysis of strategies for quality control and describes a simple representation of the components of the quality in crowdsourced GI. In this framework, and based on the research carried out in Criscuolo et al. (2014), we also introduce a methodology for quality assessment, based on the given representation, which goes beyond the limitations of previous methods in the literature defined for a specific purpose, being able to deal with many quality features, GI categories, and types of application. The method is designed as a decision making approach, so flexible as to take into account the purpose of GI analysis, and so transparent as to make explicit the criteria driving to quality evaluation, namely the quality features (e.g. the credibility of the volunteers, or the accuracy of the spatial features, etc.) and their relevance.