Brussels, Belgium Belgium
Dec. 10, 2012 to Dec. 10, 2012
In recent years, social media users are voluntarily making large volume of personal data available on the social networks. Such data (e.g., professional associations) can create opportunities for users to strengthen their social and professional ties. However, the same data can also be used against the user for viral marketing and other unsolicited purposes. The invasion of privacy occurs due to privacy unawareness and carelessness of making information publicly available. In this paper, we perform a large-scale inference study in three of the currently most popular social networks: Foursquare, Google+ and Twitter. Our work focuses on inferring a user's home location, which may be a private attribute, for many users. We analyze whether a simple method can be used to infer the user home location using publicly available attributes and also the geographic information associated with locatable friends. We find that it is possible to infer the user home city with a high accuracy, around 67%, 72% and 82% of the cases in Foursquare, Google+ and Twitter, respectively. We also apply a finer-grained inference that reveals the geographic coordinates of the residence of a selected group of users in our datasets, achieving approximately up to 60% of accuracy within a radius of six kilometers.
Cities and towns, Twitter, Privacy, Education, Employment, Accuracy, Twitter, Location, Privacy, Social Networks, Location Inference, Foursquare, Google+
Tatiana Pontes, Gabriel Magno, Marisa Vasconcelos, Aditi Gupta, Jussara Almeida, Ponnurangam Kumaraguru, Virgilio Almeida, "Beware of What You Share: Inferring Home Location in Social Networks", ICDMW, 2012, 2013 IEEE 13th International Conference on Data Mining Workshops, 2013 IEEE 13th International Conference on Data Mining Workshops 2012, pp. 571-578, doi:10.1109/ICDMW.2012.106