Users key locations in online social networks: identification and applications

Abstract

Ubiquitous Internet connectivity enables users to update their Online Social Network profile from any location and at any point in time. These, often geo-tagged, data can be used to provide valuable information to closely located users, both in real time and in aggregated form. However, despite the fact that users publish geo-tagged information, only a small number implicitly reports their base location in their Online Social Network profile. In this paper, we present a simple yet effective methodology for identifying a user’s Key locations, namely her Home and Work places. We evaluate our methodology with Twitter datasets collected from the country of Netherlands, city of London and Los Angeles county. Furthermore, we combine Twitter and LinkedIn information to construct a Work location dataset and evaluate our methodology. Results show that our proposed methodology not only outperforms state-of-the-art methods by at least 30 % in terms of accuracy, but also cuts the detection radius at least at half the distance from other methods. To illustrate the applicability of our methodology and motivate further research in location-based social network analysis, we provide an initial evaluation of three such approaches, namely (1) Twitter user mobility patterns, (2) Ego network formulation, and (3) Key location tweet sentiment analysis.