ISPRS Annals of the Photogrammetry, Remote Sensing and Spatial Information Sciences
Publications Copernicus
Articles | Volume II-3/W5
20 Aug 2015
 | 20 Aug 2015


F. O. Ostermann, H. Huang, G. Andrienko, N. Andrienko, C. Capineri, K. Farkas, and R. S. Purves

Keywords: User-generated Geographic Content, Volunteered Geographic Information, Geo-social Media, Semantic Similarity, Geographic Places

Abstract. Increasing availability of Geo-Social Media (e.g. Facebook, Foursquare and Flickr) has led to the accumulation of large volumes of social media data. These data, especially geotagged ones, contain information about perception of and experiences in various environments. Harnessing these data can be used to provide a better understanding of the semantics of places. We are interested in the similarities or differences between different Geo-Social Media in the description of places. This extended abstract presents the results of a first step towards a more in-depth study of semantic similarity of places. Particularly, we took places extracted through spatio-temporal clustering from one data source (Twitter) and examined whether their structure is reflected semantically in another data set (Flickr). Based on that, we analyse how the semantic similarity between places varies over space and scale, and how Tobler's first law of geography holds with regards to scale and places.