PLOS One
Published March 13, 2019 (News release from PLOS)
Researchers trained a computer model to search Google Street View’s images of Ottawa properties from 2007 to 2016 for changes that could indicate gentrification – for example, new fences, repainting, or new windows. The results were almost as accurate at detecting gentrification indicators as a human investigator. The information obtained could support measures for urban planning, social justice, and addressing inequality.
Lead author: Michael Sawada, University of Ottawa – msawada@uottawa.ca