Last Tuesday, René Westerholt gave a talk about the assessment of spatial autocorrelation within social media data. It was part of the regular colloquium of the Institute for Geoinformatics and Remote Sensing at Osnabrück University.
The talk provided an overview of challenges that occur due to the spatially and semantically mixed nature of social media data. The second part of the talk provided insight into current research on that topic that is conducted in the GIScience group here at Heidelberg University.
e.g. compare:
Westerholt, R., Resch, B., Zipf, A. (2015): A local scale-sensitive indicator of spatial autocorrelation for assessing high- and low-value clusters in multi-scale datasets. International Journal of Geographical Information Science (IJGIS), 29 (5), pp. 868-887, Taylor & Francis. DOI: 10.1080/13658816.2014.1002499.