Preserving the User’s Privacy in Social Networking Sites

Abstract

Abstract. In the last years, social networking sites (SNSs) have enjoyed an undeniable success. Those web platforms have huge quantities of active users sharing lots of information everyday. Usually, user-generated content may be almost innocuous, however, some studies have shown that it may also contain very sensitive personal data. This situation may pose a serious privacy threat to the users due to the fact that third parties can gather and exploit that knowledge for their own benefit. There are some proposals in the literature that try to address this situation. Nevertheless, they fail to provide a practical solution capable of working with well-known SNSs. In this paper, we propose a new scheme that fills this gap. More specifically, we present a privacy-preserving system that enables the users to decide which individuals (e.g., other users, third parties or even the SNS itself) can access to their user profiles. We have implemented our scheme to be used by Facebook users. We have run some tests with our prototype and the results show that the added overhead is affordable.

REMOVED: Enforcing transparent access to private content in social networks by means of automatic sanitization

“…Even though these privacy-preserving mechanisms represent a certain improvement with respect to the user's privacy, they have been criticized in the literature due to the following three essential problems: (i) A C C E P T E D M A N U S C R I P T these privacy settings are generally not sufficiently understood by the average users who seldom change the default configuration that generally makes most of the user information public (Bilton, 2010;Stern & Kumar, 2014;Van Eecke & Truyens, 2010); (ii) users are not informed about the privacy risks that their published data may cause and, therefore, they may find difficulties in defining effective privacy settings over their data (Wang, Nepali, & Nikolai, 2014); and (iii) these privacy settings, in any case, do not prevent social network operators from gathering sensitive user data and exploiting it to obtain economic benefits from advertisers or other entities (Crimes, 2012;Viejo, Castellà-Roca, & Rufián, 2013;Wilson, 2011). Those problems have been already identified in the literature.…”

“…Works such as (Becker & Chen, 2009;Talukder, Ouzzani, Elmagarmid, Elmeleegy, & Yakout, 2010;Wang, et al, 2014) have provided solutions that measure the users' privacy exposure by means of their published profile attributes (e.g., address, political views, religious views, etc.) and warn them when their exposure level is too high; also, schemes such as (Viejo, et al, 2013) or (Conti, Hasani, & Crispo, 2011) have been designed to protect the sensitive user data from unauthorized entities, which include social network operators. Nevertheless, despite the efforts of the scientific community, there are two related issues that still have room for improvement: (i) there is not an effective mechanism that enables users to measure the degree of sensitivity of their textual publications (i.e., how dangerous is a certain tweet/timeline post from the privacy point of view?…”

Privacy-driven access control in social networks by means of automatic semantic annotation

“…As a result, any information related to the publisher's sexual life will be sanitized from any publication accessed by friends. In contrast, the rule 6 permits family friends to know that the publisher is homosexual but nothing more detailed.…”

“…As a result, the monitor of our system will prepare a sanitized version of a message (shown in figure 11) for followers, which hides the specific details of the publisher's condition. Likewise, according to rule 6 , the access level for registered users is ill health, thus producing the sanitized message shown in figure 12. Finally, according to the first four rules (i.e., rule 1 to rule 4 ), the clinicians/researchers can access all the details of diseases of the publisher.…”

“…For this reason, most of the OSN friends are considered untrustworthy to 1 share sensitive information. On the other hand, according to the European Union Agency for Fundamental Rights & Council of Europe [5], this sensitive information needs to be protected from untrusted third parties, because it can be exploited by such parties for their own benefit [6].…”