Online auction networks, such as eBay, often use
reputation-based systems to help users assess each other's honesty and
integrity. Fraudsters, however, can collude with accomplices to
accumulate bogus positive feedbacks to manipulate reputation systems.
In this talk, we model an online auction network with fraudsters as a
random network model with hidden communities (i.e., fraudsters and
associated accomplices), and use the maximum likelihood detection framework
for detecting fraudsters. Two highly-accurate detection algorithms
will be presented.