MIT prof sees major flaws in computerized marking programs

May 5, 2014

A former MIT professor and language researcher has spoken out against computerized essay-marking platforms. Les Perelman, who was formerly Director of Undergraduate Writing at MIT, said that computer essay-marking systems are bound to become more common due to desires to cut costs. However, he says, the software is simply not good at its job. Perelman has developed a program he calls the Basic Automatic BS Essay Language Generator (BABEL) to prove his hypothesis. The program generates essays based on a handful of key words. The essays produced by BABEL are grammatically correct and structurally sound, but are nevertheless meaningless—“complete, incoherent nonsense” Perelman says. Nevertheless, these essays received top grades from the marking software. Perelman says that essay-grading programs “are not measuring any of the real constructs that have to do with writing”: they cannot discern meaning, nor can they check facts. CBC | The Chronicle of Higher Education