A reinforcement learning approach to adaptive remediation in online training

Advances in artificial intelligence (AI) and machine learning can be leveraged to tailor training based on the goals, learning needs, and preferences of learners. A key component of adaptive training systems is tutorial planning, which controls how scaffolding is structured and delivered to learners to create dynamically personalized learning experiences. The goal of this study was to induce data-driven policies for tutorial planning using reinforcement learning (RL) to provide adaptive scaffolding based on the Interactive, Constructive, Active, Passive framework for cognitive engagement. We describe a dataset that was collected to induce RL-based scaffolding policies, and we present the results of our policy analyses. Results showed that the best performing policies optimized learning gains by inducing an adaptive fading approach in which learners received less cognitively engaging forms of remediation as they advanced through the training course. This policy was consistent with preliminary analyses that showed constructive remediation became less effective as learners progressed through the training session. Results also showed that learners’ prior knowledge impacted the type of scaffold that was recommended, thus showing evidence of an aptitude–treatment interaction. We conclude with a discussion of how AI-based training can be leveraged to enhance training effectiveness as well as directions for future research.

Student Clustering Based on Learning Behavior Data in the Intelligent Tutoring System

The idea of clustering students according to their online learning behavior has the potential of providing more adaptive scaffolding by the intelligent tutoring system itself or by a human teacher. With the aim of identifying student groups who would benefit from the same intervention in AC-ware Tutor, this research examined online learning behavior using 8 tracking variables: the total number of content pages seen in the learning process; the total number of concepts; the total online score; the total time spent online; the total number of logins; the stereotype after the initial test, the final stereotype, and the mean stereotype variability. The previous measures were used in a four-step analysis that consisted of data preprocessing, dimensionality reduction, the clustering, and the analysis of a posttest performance on a content proficiency exam. The results were also used to construct the decision tree in order to get a human-readable description of student clusters.

A Problem Solving Using Intelligent Social Network

Nowadays, students are becoming familiar with the computer technology at a very early age. Moreover, the wide availability of the internet gives a new perspective to distance education making e-learning environments crucial to the future of education. Intelligent tutoring systems (ITSs) provide sophisticated tutoring systems using artificial intelligence techniques in computer programs to facilitate instruction. ITSs support problem solving by providing adaptive scaffolding in terms of feedback, guidance, problem selection and other types of help. Little attention has been devoted thus far to the difference between novices and advanced students in learning from social network and ITSs. Furthermore, students can benefit from collaboration in social network environments especially when this is combined with an ITS.

The Design of Disciplinarily-Integrated Games as Multirepresentational Systems

Disciplinarily-integrated games represent a generalizable genre and template for designing games to support science learning with a focus on bridging across formal and phenomenological representations of core science relationships (Clark, Sengupta, Brady, Martinez-Garza, and Killingsworth, 2015; Clark, Sengupta, & Virk, 2016; Sengupta & Clark, 2016). By definition, disciplinarily-integrated games (DIGs) are therefore multirepresentational systems with the affordances and challenges associated with that medium. The current paper analyzes the DIG structure through the focal parameters framed by the DeFT framework (Ainsworth, 2006) to synthesize effective design considerations for DIGs in terms of the specific design and intended functions of the representations themselves as well as the overarching environment and activity structures. The authors leverage the literatures on embodied cognition, adaptive scaffolding, representations in science education, and learning from dynamic visualizations to address the challenges, tradeoffs, and questions highlighted by the framework. They apply these research-derived design considerations to an existing DIG (SURGE Symbolic) and to hypothetical examples of other DIGs in other domains to explore generalizability of the design considerations and the genre.

An adaptive scaffolding e-learning system for middle school students’ physics learning

<p>This study presents a framework that utilizes cognitive and motivational aspects of learning to design an adaptive scaffolding e-learning system. It addresses scaffolding processes and conditions for designing adaptive scaffolds. The features and effectiveness of this adaptive scaffolding e-learning system are discussed and evaluated. An experiment was conducted within the domain of velocity and acceleration. The results revealed that the adaptive scaffolding system appeals to students and promotes improved performances and motivational outcomes. Specifically, the results suggested that learners with lower levels of knowledge who possessed extrinsic motivation benefited the most from the adaptive scaffolds. The implications of the design guidelines for developing personalized, adaptive scaffolding e-learning systems are discussed, and future research directions are recommended.</p>