Improving volunteered geographic information quality using a tag recommender system: The case of OpenStreetMap

Abstract

Studies have analyzed the quality of volunteered geographic information (VGI) datasets, assessing the positional accuracy of features and the semantic accuracy of the attributes. While it has been shown that VGI can, in some contexts, reach a high positional accuracy, these studies have also highlighted a large spatial heterogeneity in positional accuracy and completeness, but also concerning the semantics of the objects. Such high semantic heterogeneity of VGI datasets becomes a significant obstacle to a number of possible uses that could be made of the data. This paper proposes an approach for both improving the semantic quality and reducing the semantic heterogeneity of VGI datasets. The improvement of the semantic quality is achieved by using a tag recommender system, called OSMantic, which automatically suggests relevant tags to contributors during the editing process. Such an approach helps contributors find the most appropriate tags for a given object, hence reducing the overall dataset semantic heterogeneity. The approach was implemented into a plugin for the Java OpenStreetMap editor (JOSM) and different examples illustrate how this plugin can be used to improve the quality of VGI data. This plugin has been tested by OSM contributors and evaluated using an online questionnaire. Results of the evaluation suggest a high level of satisfaction from users and are discussed.