Fraud Detection Paper Appeared at CCS 2005
Posted 12 November 2005
The paper An auctioning reputation system based on anomaly
detection, co-authored by Shai Rubin, Mihai Christodorescu,
Vinod Ganapathy, Jonathon T. Giffin, Louis Kruger, Hao Wang, and
Nicholas Kidd, appeared at the ACM Conference on Computer and
Communications Security (CCS 2005). The conference was held November
7–11 in Alexandria, Virginia.
The paper, presented at the conference by Shai Rubin, argued that existing reputation systems used by online auction houses do not address the concern of a buyer shopping for commodities—finding a good bargain. These reputation systems do not provide information on practices adopted by sellers to ensure profitable auctions. The authors developed a reputation system based on anomaly detection to help buyers identify sellers whose auctions seem price-inflated. The reputation system is based on statistical metrics related to price inflation. These statistical models were combined with anomaly detection techniques to identify the set of suspicious sellers. The authors evaluated their reputation system on 604 high-volume sellers who posted 37,525 auctions on eBay. The reputation system developed in the paper automatically pinpointed sellers whose auctions contained potential shill bidders. Manual analysis of these seller's auctions revealed that winning bids were at about the item's market values, and undercut a buyer's ability to find a bargain, thus demonstrating the effectiveness of the reputation system.
