An Automated Approach to Identifying Corporate Editing

Abstract

OpenStreetMap (OSM) is the world’s largest peer-produced geospatial project. As a freely-editable open map of the world to which anyone may contribute or make use of, the dynamics and motivations of its contributors have been the object of significant scholarship. A growing phenomena in the OSM community is the increasing contributions of paid editing teams hired by tech corporations, such as, Microsoft, Apple, and Facebook. Though corporations have long supported OSM in various ways, the recent growth of teams of paid editors raises challenges to the community’s norms and policies, which are historically oriented around contributions by individual volunteer, making it hard to track the contribution of paid editors. This research addresses a fundamental problem in approaching these concerns: understanding the scale and character of corporate editing in OSM. We use machine-learning to improve upon prior approaches to estimating this phenomena, contributing both a novel methodology as well a more robust understanding of the latest corporate editing behavior in OSM.