Showing papers by "Jarrod Trevathan published in 2019"
TL;DR: Experimental results show that the algorithm is able to potentially detect colluding shill bidders and comparative analysis on simulated auction datasets shows that the proposed algorithm performs better than two existing published approaches.
16 citations
TL;DR: This paper presents the Auction Data Collector to automatically acquire vital eBay auction data required for shill bidding detection and the BidderLinker Algorithm, which links bidders across multiple auctions they have bid on with the same seller by utilising the 30-Day Bid Summary.
Abstract: Shill bidding is a fraudulent act whereby a seller places bids on his/her own auction to drive up the final price for the winning bidder. eBay currently masks bidder usernames in auctions and has adopted a 30-Day Bid Summary that restricts the availability of data on the previous auctions a bidder has participated in. This has made shill bidding detection more difficult, as tracking users across multiple auctions is essential in identifying shills. This paper presents two pieces of software to aid in the detection of shill bidding in light of masked bidder usernames and limited bid history data. First, we propose the Auction Data Collector to automatically acquire vital eBay auction data required for shill bidding detection. Second, we propose the BidderLinker Algorithm, which links bidders across multiple auctions they have bid on with the same seller by utilising the 30-Day Bid Summary. The 30-Day Bid Summary provides information about bidders that can be used to uniquely identify them, even if their usernames are masked. This allows for shill detection algorithms to be applied to the data gathered from multiple auctions with the same seller, where previously this was not possible.
4 citations