Applying Bayesian Network to Assess the Levels of Skills Mastering in Adaptive Dynamic OER-Systems
Educational models and instruction models are changing dramatically. Education is moving from the Open Educational Resources (OER) concept to the Open Educational Practices (OEP) concept. An on-demand training is becoming the norm. So, there is no option
- PDF / 6,871,367 Bytes
- 27 Pages / 439.37 x 666.142 pts Page_size
- 29 Downloads / 188 Views
,
Volga State University of Technology, Yoshkar-Ola, Mari El Republic, Russian Federation {NehaevIN,ManukyancSV}@volgatech.net, [email protected], [email protected]
Abstract. Educational models and instruction models are changing dramatically. Education is moving from the Open Educational Resources (OER) concept to the Open Educational Practices (OEP) concept. An on-demand training is becoming the norm. So, there is no option but to help the students in reaching their goals. Therefore, the diagnostic models and models for assessing the levels of skills mastering must change, they must be as flexible and diverse as the learning paths and must be data-driven. In this paper, we consider the model of the levels of skills mastering. We propose the algorithm of the initial constructing of Bayesian Cognitive Network (BCN) based on task solving data. It is shown that the BCN encapsulates the important properties of the cognitive process of mastering skills and is data-driven. The algorithm is applied to model BCN for MOOC and AI-practicum. Examples of assessing levels of mastering skills are considered. We propose some ways to predict future results and to adjust current assessment results. The issues of applying and developing models are discussed #CoMeSySo2020. Keywords: Adaptive E-Learning
Diagnostic model Bayesian network
1 Introduction 1.1
Learning Models: Base Tendencies
The world is continue taking steps towards the synergy of educational technologies, online technologies and technologies of machine learning and AI. Open online courses are becoming nowadays the modern textbooks which, compared to the traditional books, enable students to be more productive in content mastering. On the one hand, components of online courses allow students to interact not only with the course content but also with the other students and the course staff. On the other, the course may provide different learning paths to master its content. These paths may vary not only in content but, as well, in the level of difficulty, type of the content, sequence of content components of different types (video-audio, text, graphic, examples, quizzes, assignments etc.) and of different difficulty levels (High, Intermediate, Low) [1]. © The Editor(s) (if applicable) and The Author(s), under exclusive license to Springer Nature Switzerland AG 2020 R. Silhavy et al. (Eds.): CoMeSySo 2020, AISC 1294, pp. 1090–1116, 2020. https://doi.org/10.1007/978-3-030-63322-6_94
Applying Bayesian Network to Assess the Levels
1091
One more thing that makes online-courses and online-learning more effective is learning statistics. Learning statistics allows not only to create learning styles and level of knowledge models but also to create the models that enable learning outcomes prediction and learning assistance planning [2, 3]. Adding to that, technology-enhanced learning (TEL) brings online learning to a new level of teaching and learning assistance. Learning platforms and online-courses implements recommender systems that analyze not only level
Data Loading...
