Online auctions are gaining tremendous popularity in re-
cent years. Although providing unprecedent opportunities, online auc-
tion sites become an attractive environment for fraud. The expansion of
the share of online auctions in the world trade causes exponential growth
of theft and deception associated with this medium. Participants of on-
line auctions agree that trustworthy reputation systems are an important
factor in fighting dishonest and malicious users. Unfortunately, popular
auction sites use only very simple reputation estimation schemes that
utilize feedbacks issued reciprocally by users after terminated auctions.
Such systems can be easily deceived and do not offer sufficient protection
against organized fraud.

In this paper we present a novel density-based reputation measure. The
new reputation measure uses the topology of seller-buyer connections to
derive knowledge about trustworthy sellers. We mine the data on past
transactions to discover clusters of connected sellers and for each seller we
measure the density of the seller’s neighborhood. We use these clusters
both for scoring the reputation of individual sellers, and to assist buyers
in informed decision making by generating automatic recommendations.
We perform many experiments on the body of real-world data acquired
from a leading Polish provider of online auctions to examine the new
measure in detail.