The usage of OSNs introduces numerous security
and privacy threats. For instance, as a user
needs to interact with other users through an
OSN service provider, its activities and uploaded
data can be tracked and stored by the OSN service
provider. These data (photos, articles, public
posts, private messages, etc.) may be leaked
to a third party without the user’s explicit authorization,
even when the user regards some of
these as confidential. Moreover, Sybil attacks are
very common in OSNs, as a user can register
multiple fake accounts maliciously. These fake
accounts can perform various malicious activities
including spamming, obtaining privacy contact
lists, misleading crowd-sourcing results, and so
on. Besides those, Gao et al. [16] list several
other attacks such as re-identification and deanonymization
of anonymized OSN data, fetching
personal data through untrusted third-party
applications, cross-site profile cloning, social
spamming, and phishing. Due to space limitation,
this survey mainly focuses on malicious
behavior in OSNs, including spam and Sybil
attacks.