|
A Computer Model to Detect Fraudulent Internet AuctionsDr John Gochenouer, Dr Gerald Toland. This empirical study tracked 299 new participants in online auctions at eBay. All of the selling and buying activities of these persons were tracked and recorded over a seven month period. Images of the products were compared with images of products that were alleged to have been previously sold. The feedback that they received from persons they transacted with was studied and analysed. This resulted in thousands of data items to be interpreted.
Over seven months, many patterns of behavior emerged including the practice of shilling. Shilling is the anonymous bidding on auctions by the seller and is not allowed for both ethical and legal reasons. Our analysis reaches the conclusion that logic may be codified into computer models to detect this practice. While eBay claims that a miniscule number of auctions are fraudulent, our modest preliminary study suggests that the number of illegal auctions is exponentially higher. The greatest asset that eBay has is the belief by millions of people that they are engaged in fair transactions. Any threat to this asset should be acted on by the company. We offer a stepwise model that will minimize violations of privacy, wrongfully contacting innocent buyers, and the need for massive data mining. Presenters  Dr John Gochenouer
(United States)
Professor Business and Public Affairs Department Southwest Minnesota State University Dr. John Gochenouer has a BS/MBA from the University of Maryland and then served as a researcher at the National Cancer Institute for three years. He became the Information Systems Director at the multi-billion dollar Harris Corporation. In 1985 he received a Ph.D. from Florida Tech and has since authored many academic articles on the internet. In 1990, he participated in the launch of an MBA program in Kingston Jamaica and in 2000 became the Graduate Management Program Director at SMSU. He implemented a new MBA program in 2002 that today is accredited by the Higher Learning Commission.
Dr Gerald Toland
(United States)
Southwest Minnesota State University
Keywords
(30 min. Conference Paper,
English)
|