A multi-objective differential evolution approach for the question selection problem

Examinations are important tools for assessing student performance. They are commonly used as a metric to determine the quality of the students. However, examination question paper composition is a multi-constraint concurrent optimization problem. Question selection plays a key role in question paper composition system. Question selection is handled in traditional systems by using a specified question paper format containing a listing of weightages to be allotted to each unit/module of the syllabus. They do not consider other constraints such as total time duration for completion of the paper, total number of questions, question types, knowledge points, difficulty level of questions etc,. In this paper we have proposed an innovative evolutionary approach that handles multi-constraints while generating question papers from a very large question bank. The proposed Multi-objective Differential Evolution Approach (MDEA) has its advantage of simple structure, ease of use, better computational speed and good robustness. It is identified to be more suitable for combinatorial problems as compared to the generally used genetic algorithm. Experimental results indicate that the proposed approach is efficient and effective in generating near-optimal or optimal question papers that satisfy the specified requirements.