Can social media data be useful in spatial modelling? A case study of ‘museum Tweets’ and visitor flows

This paper explores the potential of volunteered geographical information from social media to inform geographical models of behavior. Based on a case study of museums in Yorkshire, we created a spatial interaction model of visitors to 15 museums from 179 administrative zones to test this potential. Instead of relying on limited official data on the magnitude of flows from different attractions we used volunteered geographic information’ (VGI) to calibrate the model. The method represents the potential of VGI for applications beyond descriptive statistics and visuals and highlights potential uses of georeferenced social media data for geographic models. The main input dataset comprised geo-tagged messages harvested using the Twitter Streaming Application Programming Interface (API). We successfully calibrated the distance decay parameter of the model and conclude that social media data have great potential for aiding models of spatial behavior. However, we also caution that there are dangers associated with the use of social media data. Researchers should weigh up the wider costs and benefits of harnessing such ‘big data’ before blindly harnessing this low quality, high volume resource. Our case study also serves as the basis for discussion of the ethics surrounding the use of privately harvested VGI by publicly funded academics.