Creating questions on an exam is a complex process, because this process requires knowledge and takes a long time to make. The process of creating questions can be done more easily, quickly, and structured with the Automatic Question Generator (AQG) system. This application uses the Text Matching Method to find keywords in a paragraph, where the Expected Answer Type (EAT) method would identify these keywords. The EAT method helped to identify the type of answers in a paragraph therefore the type of questions would be recognized. The types of questions used are 5W + 1H consisting of Who, Where, When, Why, What, How, and How Many.The next method is the Template Based Method which played a role in compiling the question sentence based on the pre-registered template. The questions were produced using the Revised Bloom's Taxonomy concept, where these questions consisted of categories (1) remembering; (2) understand; (3) apply; (4) analyze; (5) evaluate; and (6) create. The trial result in 14 learning materials, showed that the application could generate 826 questions with an average level of accuracy of 89%.
Copyrights © 2021