A University of Regina Computer Science Professor, Samira Sadaoui, believes online auction scams are taking place more often than previously thought. Me too!
But unlike me, she isn’t just talking about it. She’s doing something interesting about it. She and a research group are building software application to find fraudulent behaviour in online auctions as it is happening.
If you have ever purchased something on eBay, you understand the disappointment of a bidding war. Many times, the buyer you think you’re battling in an online auction might not really want the product and is only seeking to drive up the cost.
I’ve been frustrated for years by this issue. I buy domain names a lot in auctions on eBay and GoDaddy and I know it’s not fun how some bidders seem determined to buy the product at any price, beat up the price and then vanish close to the end of the sale, leaving an impression that they never wanted to buy the product even for a dollar.
Just like an in-person auction can be subject to abuse if not well monitored, the same inappropriate behavior often surface online. For in-person auctions, the cheat bidder might appear early on, beat up the price and disappear when it’s becoming clear that the auctioneer would soon bang the gavel to announce the emergence of a winner.
Online, the cheat bidder could be a real person or an automated program. It really doesn’t matter. The pattern is consistent, and should be pretty easy to control if the bidding platforms cared. But since it’s all to their profit, they simply choose to condone it – that is, if they aren’t actively masterminding it.
As online auction grows in popularity, some folks just seem more determined than ever to game everyone else. For those of us who are legitimate bidders, it’s really pricey.
There are three kinds of online auction fraud: pre-auction fraud, post-auction fraud and in-auction fraud. Sadaoui says her research focuses on the last one and developing powerful software to detect the pattern and curb it in real time.
They already have one prototype for their detector. Once it’s put to vigorous testing and stabilized, it would be improved and patented, and then possibly sold to online auction sites like eBay and GoDadddy.
The auction websites might use the software to discover scammers as well as suspend their accounts. The software would likely establish patterns, monitor IP addresses, cookies, and limit or ban the access of suspected or established fraudulent users.
While this is plausible, I think the research team has a huge task ahead of them. A high level of artificial intelligence will need to be programmed into the software to make it effective. Again, they would have to treat it like all software – not a done-and-forget project but a constantly evolving one as the online landscape changes daily with scammers getting smarter by the second.