Online Social Networks (OSNs) allow users to build connections, establish relationships, and exchange information over the Internet. At the same time, OSNs hold an abundant amount of data that are insufficiently protected by default privacy preferences. Moreover, privacy preferences are by default hard to use and do not correctly reflect the intentions of users, which may lead to leakage of information to a broader audience.
To analyze the leakage of information in OSNs, such as Facebook, Giuseppe Cascavilla and his colleagues propose to use Open Source InTelligence (OSINT) techniques to extract and infer information from publicly available data sources. The aim of their study is twofold:
- To reconstruct personal and supposed hidden friends list;
- To infer personal private information like work, education, hometown, current city of a victim user from his social network.
The authors built a system, named OSSINT (Open Source Social Network Intelligence), which retrieves the common friends between the owner of the friends list and all the IDs from its friends found list. OSSINT can be used to predict multiple-hop friendships such as friends-of-friends. Besides, OSSINT manages to reconstruct the friendship graph of a victim user along with the importance weight of friends.
Their system demonstrated that Open Source Intelligence allows retrieving a significant amount of information that users consider, set, and believe is kept private to any prying eyes or third parties. The authors were capable of rebuilding the friendship graph of a victim user and evaluate the weight of each friendship based on the number of shared friends. Finally, they were able to rebuild other private attributes such as personal information, demonstrating that it was possible to infer real-life information supposed being private.
Cite: Cascavilla, G., Beato, F., Burattin, A., Conti, M. and Mancini, L. V.(2018). OSSINT – Open Source Social Network Intelligence: An efficient and effective way to uncover “private” information in OSN profiles