Collaborating with Bots and Automation on OpenStreetMap

Abstract

OpenStreetMap (OSM) is a large online community where users collaborate to map the world. In addition to manual edits, the OSM mapping database is regularly modified by bots and automated edits. In this paper, we seek to better understand how people and bots interact and conflict with each other. We start by analysing over 15 years of mailing list discussions related to bots and automated edits. From this data, we uncover five themes, including how automation results in power differentials between users and how community ideals of consensus clash with the realities of bot use. Subsequently, we surveyed OSM contributors on their experiences with bots and automated edits. We present findings about the current escalation and review mechanisms, as well as the lack of appropriate tools for evaluating and discussing bots. We discuss how OSM and similar communities could use these findings to better support collaboration between humans and bots.