Characterizing Volunteered Geographic Information using Fuzzy Clustering

This paper demonstrates the use of fuzzy clustering to characterize Volunteered Geographic Information (VGI). We argue that classifying small areas based on variables related to the amount, type, and currency of VGI can provide a more nuanced understanding of the content. We present a classification of 2011 UK Census Output Areas in Leicestershire (UK) based on content of OpenStreetMap, using a fuzzy c-means clustering algorithm, and we compare the resulting classification with a ‘standard’ socio-economic geodemographic classification.