Sign In

Communications of the ACM

ACM TechNews

Hoax-Detecting Software Spots Fake Papers

View as: Print Mobile App Share:
A robot typing.

A new tool to combat nonsense paper submissions uses a statistical technique to automatically detect papers created by software programs.

Credit: Andrey Voskressenskiy/iStock

In 2005, three Ph.D. students at the Massachusetts Institute of Technology developed SCIgen, a program to generate nonsense computer science papers as a way of unmasking scientific conferences and journals that were failing to carry out genuine peer review of submissions. Since then, SCIgen, and others like it, have become sources of embarrassment for major academic publishers.

Germany-based publisher Springer and the Institute of Electrical and Electronic Engineers (IEEE), for example, have had to retract 122 nonsense papers.

This month, Springer announced the creation a new tool meant to combat nonsense paper submissions. Called SciDetect, the tool was developed in partnership with Cyril Labbe, a computer scientist at Grenoble's Joseph Fourier University. The tool uses a statistical technique similar to those used by email spam filters to automatically detect papers created with SCIgen and similar programs.

Springer says it developed SciGen in response to a recent scandal involving Chinese students and academics padding their resumes with articles generated by SCIgen and similar programs. However, critics say the technology provides publishers with an excuse to not employ peer-reviewers and will mark a return to the old status quo SCIgen was created to criticize.

From Science Insider
View Full Article


Abstracts Copyright © 2015 Information Inc., Bethesda, Maryland, USA


No entries found

Sign In for Full Access
» Forgot Password? » Create an ACM Web Account