Cluster-Based Analysis and Recommendation of Sellers in Online Auctions

The expansion of the share of online auctions in electronic trade causes exponential growth of theft and deception associated with this retail channel. Trustworthy reputation systems are a crucial factor in fighting dishonest and malicious users. Unfortunately, popular online auction sites use only simple reputation systems that are easy to deceive, thus offering users little protection against organized fraud. In this paper we present a new reputation measure that is based on the notion of the density of sellers. Our measure uses the topology of connections between sellers and buyers to derive knowledge about trustworthy sellers. We mine the data on past transactions to discover clusters of interconnected sellers, and for each seller we measure the density of the seller’s neighborhood. We use discovered clusters both for scoring the reputation of individual sellers, and to assist buyers in informed decision making by generating automatic recommendations. We perform experiments on data acquired from a leading Polish provider of online auctions to examine the properties of discovered clusters. The results of conducted experiments validate the assumptions behind the density reputation measure and provide an interesting insight into clusters of dense sellers.