Sign In

Communications of the ACM

ACM TechNews

Automatically Grading Programming Homework


A humanoid robot, representing automation of the programming review. A

A new software program already in use at MIT finds inaccuracies in programs created by students, and points out the most expeditious ways to fix those issues.

Credit: teachthought.com

Researchers at the Massachusetts Institute of Technology (MIT) Computer Science and Artificial Intelligence Laboratory and Microsoft Research have created software that points out inaccuracies in student programming and offers suggestions for improvement.

Already in use in MIT classes, the software could help automate grading, which is one of the largest challenges facing massive open online courses (MOOCs).

"One challenge, when [teaching assistants (TAs)] grade these assignments, is that there are many different ways to solve the same problem," says MIT graduate student Rishabh Singh. "For a TA, it can be quite hard to figure out what type of solution the student is trying to do and what’s wrong with it." Regardless of the programmer’s approach, the software identifies the minimum number of corrections necessary to fix a program.

The researchers are assessing the software's potential for grading homework assignments in programming MOOCs, considering variations that would reveal the location and nature of errors with varying levels of specificity to create appropriate learning opportunities for students.

From MIT News
View Full Article

 

Abstracts Copyright © 2013 Information Inc., Bethesda, Maryland, USA


 

No entries found