(KDKA) PITTSBURGH Computer scientists at Carnegie Mellon University are developing a new way to fight online auction fraud.

Right now, the most assurance bidders get is the feedback area, where previous buyers post ratings on a seller’s reliability.

CMU Professor Christos Faloutsos says these ratings are where the so-called “fraudsters” strike.

"What fraudsters are trying to do is that they create fake alter egos or they get their friends to lie and say, 'Hey Smith123 is actually a good person because I bought XYZ'," he explained.

But Smith123 turns out to be a scam artist who takes your money and runs.

A new computer program called NetProbe, developed by Faloutsos and his colleagues, is trying to put and end to that.

According to Faloutsos, the program tracks the “fraudsters” with red dots and the accomplices who lie for them with yellow dots.

When they interact, the program finds a pattern that can be flagged as fraud immediately.

The professor hopes to eventually have a version of NetProbe that can be used by anyone, but admits that online criminals will try to find a way to beat it.

“It’s an arms race. But the hope is that our algorithms will make fraud so difficult that eventually fraudsters will get discouraged and give up," said Faloutsos.