Discovering Multiple Design Approaches in Programming Assignment Submissions
This addresses the challenge for instructors in automated programming assignment evaluation by reducing human intervention and improving flexibility.
The paper tackles the problem of automatically identifying multiple solution approaches in programming assignment submissions, enabling automated evaluation without requiring instructors to foresee all possible approaches. The method successfully detected solution approaches across multiple practical-scale datasets.
In this paper, we present a novel approach of automated evaluation of programming assignments~(AEPA) the highlight of which is that it automatically identifies multiple solution approaches to the programming question from the set of submitted solutions. Our approach does not require the instructor to foresee all the possible solution approaches and accomplishes this task with little or no human intervention. This paves the way to multiple fundamental improvements in the way automated evaluation of programming assignments is done today. We have applied our method on multiple data sets of practical scale. In all cases, our method was able to detect the solution approaches employed by the students.