文章基本信息

标题：Generation of e-Learning tests with different degree of complexity by combinatorial optimization
本地全文：下载
作者：Daniela Ivanova Borissova ; Delyan Keremedchiev
期刊名称：Je-LKS
印刷版ISSN：1826-6223
电子版ISSN：1971-8829
出版年度：2020
卷号：16
期号：2
页码：17-24
DOI：10.20368/1971-8829/1135016
出版社：Casalini Libri
摘要：The major challenge in the digital era is the management of big data. A substantial share of digitization is taken from the e-learning. In this respect, the current article deals with generation of questions for testing the acquired levels of students’ knowledge. For this purpose, an algorithm for generation of questions for tests with different level of complexity is proposed. The main stage of this algorithm is using of mathematical combinatorial optimization model. Using this model makes it possible to formulate different tasks, whose solutions determine a subset of questions that correspond to different degree of test complexity. Essential part of this model is the use of binary integer variablesto determine whether a question will be a part of the test or not. The advantage of the proposed approach is the flexibility to decrease or increase the number of questions used to compose the test preserving the required score in accordance to the particular level of test complexity. The conducted investigations over a year show, that the effect of testing can improve retention of knowledge and lead to improved end results. The applicability of the proposed algorithm along with the formulated mathematical model is demonstrated in a case study on the excerpt of questions from the web programming course. The proposed algorithm could be used for generation of tests with different degree of complexity for other learning contents.
关键词：Combinatorial optimization;e-learning;Mathematical model;Questions difficulties;Test generation