AUTOMATIC TEST ITEM GENERATION FROM KNOWLEDGE STRUCTURE

Abstract

This research designs and implements an item generation engine that can automatically create higher order thinking multiple-choice items for online tests based on knowledge maps developed by teachers. Furthermore, this study leveraged questionnaire to collect data from teachers to analyze the agreement between system and participants classification of the cognitive items generated by the algorithms designed and implemented for this study. Results indicated that there are areas where the participants agreed with the systems classification of the cognitive items and in some areas they disagree. However, the system implemented for this research might go a long way to help teachers save the time they need to spend on preparing tests and assessing their students’ understandings of the concepts they have learnt. Moreover, students will benefit from the online test system in terms of having opportunity to self-assess their knowledge at any time and getting rapid test results.