Most users have several Internet names. On Facebook or LinkedIn, for example, people usually appear with the real one. On other standard websites, like forums, people often use aliases to protect their real identities with respect to the other users, with no real privacy against the web site and the authorities. Aliases in the Dark Web are different: users expect strong identity protection. In this paper, we show that using both “open” aliases (aliases used in the standard Web) and Dark Web aliases can be dangerous per se. Indeed, we develop tools to link Dark Web to open aliases. For the first time, we perform a massive scale experiment on real scenarios. First between two Dark Web forums, then between the Dark Web forums and the standard forums. Due to a large number of possible pairs, we first reduce the search space cutting down the number of potential matches to a small set of candidates, and then on the selection of the correct alias among these candidates. We show that our methodology has excellent precision, from 87% to 94%, and recall around 80%.
A Light in the Dark Web: Linking Dark Web aliases to real internet identities / Arabnezhad, E.; la Morgia, M.; Mei, A.; Nemmi, E. N.; Stefa, J.. - 2020-:(2020), pp. 311-321. (Intervento presentato al convegno 40th IEEE International Conference on Distributed Computing Systems, ICDCS 2020 tenutosi a Singapore) [10.1109/ICDCS47774.2020.00081].
A Light in the Dark Web: Linking Dark Web aliases to real internet identities
la Morgia M.
;Mei A.;Nemmi E. N.
;Stefa J.
2020
Abstract
Most users have several Internet names. On Facebook or LinkedIn, for example, people usually appear with the real one. On other standard websites, like forums, people often use aliases to protect their real identities with respect to the other users, with no real privacy against the web site and the authorities. Aliases in the Dark Web are different: users expect strong identity protection. In this paper, we show that using both “open” aliases (aliases used in the standard Web) and Dark Web aliases can be dangerous per se. Indeed, we develop tools to link Dark Web to open aliases. For the first time, we perform a massive scale experiment on real scenarios. First between two Dark Web forums, then between the Dark Web forums and the standard forums. Due to a large number of possible pairs, we first reduce the search space cutting down the number of potential matches to a small set of candidates, and then on the selection of the correct alias among these candidates. We show that our methodology has excellent precision, from 87% to 94%, and recall around 80%.I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.