Moodle plugins for quiz generation using genetic algorithm

Abstract

This paper describes the development of Moodle plugins to generate quiz using Genetic Algorithm. Moodle is an open source learning management system. One of its features is to generate a quiz consisting random questions which is chosen form question bank. If the quiz is created randomly, it can't accommodate specific constraints. Genetic algorithm provides an opportunity to generate quiz which approximate the constraints. Our research follows previous researches done by Xiaoqin & Yin (2009) and Huang & Wang (2008), modifies them to be applied in the development of Moodle quiz generation plugins. Several attributes which are selected from previous researches, are augmented to the each question in question bank. The generated quiz is also augmented by corresponding attributes. Quiz generator needs constraints as input. Genetic algorithm will generate a quiz whose attributes approximate the constraints. The generated quiz can be used in tests after this step. At the moment, there is no option of using genetic algorithm for quiz generation in Moodle. The plugins consist of one main quiz plugin and several question type plugins. The main quiz plugin was developed so that genetic algorithm can be used in quiz generation. Meanwhile, several question type plugins were developed to provide augmented questions which quiz plugin is going to use. These plugins have been successfully implemented and integrated to Moodle.