Social network analysis of public lists of POIs
Ioannis E. Karagiannis, Avi Arampatzis, Pavlos S. Efraimidis, Giorgos Stamatelatos · 2015
In this work, we show how social network analysis can be applied to lists of points of interest (POIs) in order to extract important information about the POIs and the relations between them. More precisely, we use public lists of POIs to build the PoiGraph, a social graph of POIs, and then apply the Hyperlink-Induced Topic Search (HITS) algorithm and the Normalized Pointwise Mutual Information (NPMI) measure to estimate the user rating of each POI and the pairwise similarity between POIs, respectively. We evaluate our approach on POIs from the cities of Athens, Thessaloniki, and Rhodes. As a data source we use the corresponding publicly accessible user-specified lists of POIs of Foursquare. Our results show that for each POI the authority score obtained with the HITS algorithm is firmly correlated with the actual rating of Foursquare. Moreover, preliminary evidence shows that the NPMI-based measure gives valuable information about the pairwise similarity between POIs.