Quality Evaluation of Volunteered Geographic Information: The Case of OpenStreetMap

Background

The term volunteered geographic information (VGI) was suggested by Goodchild (2007) to represent geospatial data contributed by individuals voluntarily. Since VGI is often the most cost-effective solution, the crowd-sourced geodata have been applied in many fields such as participatory planning and spatial decision making. Moreover, VGI is the only source of geodata in some regions because of security or financial concerns. The area of humanitarian relief and crisis management is the most prominent application of VGI. Ushahidi and the Humanitarian OpenStreetMap Team (HOT) are two platforms that have had strong presence on disaster management since 2008 and 2009 respectively. Table 1 compares some VGI applications with OSM. Although OSM is not the project with the longest history, it is the oldest mapping project in which the geo-information can be applied in more than one field. The number of “registered members” of OSM is relatively small comparing to other specialized applications, but the number of “users” could be a bloated figure and does not represent “active contributors”. Jokar Arsanjani and Bakillah (2014), Mooney and Corcoran (2012a) and Yang, Fan, and Jing (2016) provide more insights into the OSM contribution patterns and user behavior. Like Wikimapia and Waze, OSM has a worldwide coverage. The difference is that OSM allows users to freely alter and redistribute its data, which is accessible through multiple servers in different formats. In contrast, Wikimapia only offers its data by a web application programming interface (API) (Neis & Zielstra, 2014), and Waze does not release data from its platform. Therefore, OSM was chosen to be the focus of this chapter. The following subsections starts with the discussion of quality concerns in VGI, introduces OSM in details and ends with a list of the spatial data quality metrics.