Thursday, February 27, 2014

Computer-generated fake papers are flooding academia

This story is all over the Internet, even making the Guardian newspaper. I'm not entirely sure I agree with the statement that fake papers are "flooding" academia, but nonetheless it is worrying that bogus papers are being published by reputable publishers like Springer and the IEEE. However, in their defense these organisations/publishers do delegate the decisions on academic quality to individual conference chairs and programme committees. It's also not true that conferences with poor or negligible standards are just restricted to China. In 2004 I had two MSc students write and submit papers to a conference taking place in Wellington, New Zealand (KES2004). The students' work was not very good and the papers they wrote were quite unexceptional and I fully expected that they would be duly rejected. I thought the rejection would be a good learning experience for the students and may encourage them to work harder in future; both were somewhat arrogant.
    I was very surprised, and slightly mortified, when both papers were accepted by the conference. My name was on these papers as co-author and knew them to be substandard. Nonetheless, I thought that attending the conference might be a good experience for the students. When they returned I was very surprised to see that the conference proceedings, published by Springer in their LNCS series, ran to two volumes of 600+ pages each. Nowhere in the proceeding's introduction did it say how many papers were submitted for review and what percentage were accepted. This conference series, run by a respected UK academic, seemed like a vanity publishing project to me.