| Autor: |
Noratikah Shamsudin, Tengku Nurulhuda Tengku Abd Rahim, Rose Hafsah Ab Rauf, Zalilah Abd Aziz |
| Rok vydání: |
2017 |
| Předmět: |
|
| Zdroj: |
2017 IEEE Conference on e-Learning, e-Management and e-Services (IC3e). |
| DOI: |
10.1109/ic3e.2017.8409231 |
| Popis: |
Manual preparation of exam questions is a challenging task for educators especially within a short time frame. It requires a lot of time and efforts in order to meet the standard quality of exam questions. This research introduces an automated exam question generator to resolve this issue in preparation of multiple choice exam questions. The generator can auto generate new exam questions set using Genetic Algorithm and covers six levels of Bloom's Taxonomy to produce high quality exam questions that can evaluate different level of learners based on Bloom's cognitive domains and the selection of chapters made by educators. The prototype with 500 sample questions has been run 50 times with different number of chapters selected for each test case. It manages to achieve 90% for the highest exam questions weightage while the average value of exam questions weightage percentage generated is 70%. The lowest exam questions weightage percentage generated is 40%. The result is affected by the smaller number of questions for each Bloom's taxonomy level in questions bank. The automated exam generator can extends to be used for any type of exam questions and it can be used for preparation of quiz or test questions. |
| Databáze: |
OpenAIRE |
| Externí odkaz: |
|
