Nature had an interesting article recently on the publication of computer-generated scientific papers:
Note that the fake papers were actually conference proceedings, and the article goes on to say that these types of papers are relatively easy to spot if people actually look for them. Apparently, some researchers at MIT were bored and created SCIgen, software to put together strings of computer-science-related text and make a paper out of them. SCIgen is freely available:
This whole concept is, to me, simultaneously funny, sad, frustrating, and spooky. Funny, well, because it just is, and this is actually the point of SCIgen. Sad because the conference organizers didn’t catch the bogusness of the abstract before it made it into the program. Frustrating because it fuels the flames of the anti-science constituency that feeds on any disparaging news about STEM fields. Spooky because someone could actually write this kind of software that generates something plausible enough to get through a cursory check. I am horrified to think that something like this could get through JGR-Space Physics. Please, as funny as it might seem, don’t waste the editorial board’s and reviewers’ time with a bogus paper; in the end it is your scientific reputation that will suffer for it and the field as a whole that will pay an indirect price for the bad press.