Work Always in Progress: Analysing Maintenance Practices in Spatial Crowd-sourced Datasets

Crowd-mapping is a form of collaborative work that empowers users to share geographic knowledge. Despite geographic information being intrinsically evolving, little research has so far gone into analysing maintenance practices in these domains. In this paper, we quantitatively capture maintenance dynamics in geographic crowd-sourced datasets, in terms of: the extent to which different maintenance actions are taking place, the type of spatial information that is being maintained, who engages in these practices and where. We apply this method to 117 countries in OpenStreetMap, one of the most successful examples of geographic crowd-sourced datasets. Furthermore, we explore what triggers maintenance, by means of an online survey to which 96 OpenStreetMap contributors took part. Our findings reveal that, although maintenance practices vary substantially from country to country in terms of how widespread they are, strong commonalities exist in terms of what metadata is being maintained, by whom, and what triggers them.