Intelligent agent for checking examination questions (IACEQ) / Nur Amalina Kamarudin

In the context of UiTM, the exam questions for each course are made by different lecturers. In preparing the final exam questions, each lecturer need to send the exam questions to the panel to decide which exam questions meet the requirement to be the final exam questions for the subject. Usually, lecturer does not compare the questions with the previous semester exam questions. Thus, the possibility of the questions being repeated again with the previous exam questions might be high. The problem is there is no intelligent agent that can measure the similarity of the questions when needed. The
purpose of this paper is to developanintelligent agent that will measure the similarity of the proposed exam questions with the previous exam questions. In the process of measuring the degree of the similarity of the questions, exact matching technique with synonym was chosen. The agent compares the proposed exam questions with the previous exam questions selected by the lecturer and generate result showing the degree of similarity between those questions. As a result, it is found out that all of the results are in range of 0.0056 to 0.1167 which is far from 1.000. This means, the proposed exam question and previous exam questions mostly are not similar. Only the same questions that are compared together produce 1.000 similarities which mean it is absolutely the same. This is proven by the result given by the system. IACEQ system can ease a lot of the lecturer task, saves their times and energy, but with some improvements. The hope is that the system can give benefit to the education field and help to reduce lecturer task.