自适应在线学习测评研究及其应用

发布时间：2018-06-27 15:36

本文选题：计算机自适应测验 + 项目反应理论　；参考：《电子科技大学》2017年硕士论文

【摘要】：随着互联网的快速发展,越来越多的学习者选择通过互联网进行在线学习,各种基于智能化和自动化的在线学习模式、方法方兴未艾,在线学习的自适应测评就是其中的一个重要方面。本文针对自适应在线学习测评理论与技术展开了深入研究,将教育学、心理学等最新成果应用到在线学习测评的研究当中,提出了针对学习者个体的自适应选题策略算法,并在此基础上实现了在线学习的自适应测评系统,提高了学习者的测评效率,为学习者高效地进行个性化在线学习能力测评提供了新的途径。论文主要进行了三个方面的工作:一、研究并设计自适应在线学习测评系统的选题策略,通过研究自适应测验的经典选题策略,分析最大信息量法、a分层法以及其改进算法的特点与局限,在经典选题策略的基础上提出了新的可靠、可行的改进选题策略,同时与传统选题策略及其改进算法从项目曝光率、题库平均曝光率、测验准确性、测验效率和测验重叠率等多个维度进行了性能比较。二、研究基于蒙特卡洛模拟的自适应测评选题策略的检验方法,对本文提出的算法进行了模拟实现,设计检验方法实验程序结构并编写检验方法程序,应用检验方法模拟选题策略测评过程,并对传统选题策略与本文提出的改进策略进行比较。三、设计并实现了自适应在线学习测评系统,基于可用性和可靠性的考虑,设计了测评系统架构,实现了测评系统各模块功能,建立了自适应在线学习测评系统的测评题库,为学习者进行在线测评提供了有效途径。论文提出的新型自适应测评算法与模式,有效降低了传统方法的项目曝光率,相对于其他分层方法提高了测验精度,在测验准确性和测验效率上都有较大提升,开发的自适应在线学习测评系统为学习者个性化学习能力的区分提供了可靠的测评手段,具有良好的应用前景和价值。
[Abstract]:With the rapid development of the Internet, more and more learners choose to learn online through the Internet. All kinds of online learning models based on intelligence and automation are in the ascendant. Adaptive evaluation of online learning is one of the most important aspects. This paper is a deep study of adaptive online learning evaluation theory and technology. In the study, the latest achievements of education and psychology are applied to the study of online learning evaluation. An adaptive selection strategy algorithm is proposed for the individual of the learners. On this basis, an adaptive evaluation system for online learning is realized, which improves the evaluation efficiency of the learners and performs the personalized online learning for the learners efficiently. Ability evaluation provides a new way. The thesis mainly carries out three aspects: first, research and design the selection strategy of adaptive online learning evaluation system, through the study of the classic selection strategy of adaptive test, the analysis of the maximum information method, the a stratification method and its improved algorithm characteristics and limitations, the basis of the classic topic selection strategy. A new reliable and feasible selection strategy is proposed. At the same time, the performance is compared with the traditional selection strategy and its improved algorithm from the project exposure rate, the average exposure rate of the question bank, the test accuracy, the test efficiency and the test overlap rate. Two, the test of the adaptive selection strategy based on the Mongol Tekalo simulation is studied. Method, the algorithm proposed in this paper is simulated, the test method experiment program structure is designed and the test method program is written. The test method is used to simulate the evaluation process of the selected topic strategy, and the traditional selection strategy is compared with the improved strategy proposed in this paper. Three, the adaptive Online learning evaluation system is designed and implemented, based on the availability of the system. Considering the nature and reliability, the architecture of evaluation system is designed, the functions of each module of the evaluation system are realized, and the evaluation question bank of the self-adaptive online learning evaluation system is set up. It provides an effective way for the learners to evaluate the online evaluation system. The new adaptive evaluation algorithm and model proposed in this paper has effectively reduced the exposure rate of the traditional methods. Compared with other stratification methods, the test accuracy is improved and the test accuracy and test efficiency have been improved greatly. The developed adaptive online learning evaluation system provides a reliable evaluation method for the differentiation of learners' individualized learning ability, and has a good application prospect and value.
【学位授予单位】：电子科技大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：G434

【参考文献】