Toward Adaptability of E-Evaluation: Transformation from Tree-Based to Graph-Based Structure

Abstract

The COVID-19 pandemic and quarantine have forced students to use distance learning. Modern information technologies have enabled global e-learning usage but also revealed a lack of personalization and adaptation in the learning process when compared to face-to-face learning. While adaptive e-learning methods exist, their practical application is slow because of the additional time and resources needed to prepare learning material and its logical adaptation. To increase e-learning materials’ usability and decrease the design complexity of automated adaptive students’ work evaluation, we propose several transformations from a competence tree-based structure to a graph-based automated e-evaluation structure. Related works were summarized to highlight existing e-evaluation structures and the need for new transformations. Competence tree-based e-evaluation structure improvements were presented to support the implementation of top-to-bottom and bottom-to-top transformations. Validation of the proposed transformation was executed by analyzing different use-cases and comparing them to the existing graph-to-tree transformation. Research results revealed that the competence tree-based learning material storage is more reusable than graph-based solutions. Competence tree-based learning material can be transformed for different purposes in graph-based e-evaluation solutions. Meanwhile, graph-based learning material transformation to tree-based structure implies material redundancy, and the competence of the tree structure cannot be restored.

Special Issue on Innovations in the Field of Cloud Computing and Education

“…Naveed et al [4] identified the service quality factors that affected students' acceptance of cloud e-learning systems and tested them in different Saudi Arabian universities. Margien ė and Ramanauskait ė [5] described techniques to facilitate learning personalization and adaptation in eLearning contexts based on transformations from a competence tree-based structure to a graph-based automated e-evaluation structure. Sein-Echaluce et al [6] found that visually representing students' cloud computing system interactions makes it possible to achieve workload homogeneity between teams and among team members.…”

Automated Transformation from Competency List to Tree: Way to Competency-Based Adaptive Knowledge E-Evaluation

“…For bigger integration of different e-learning systems, as well as the implementation of higher-level or e-learning adaptability and personalization, the transformation between different data structures are preferable. These types of transformation already exist-an automatic method of transformation from competency-based tree structure assessment data to a graph structure was proposed in 2021 [17]. In this case, competencies by hierarchy are stored in a competency tree-based structure, and a graph-based structure is used for adaptive task presentation.…”

Formative assessment methods for intelligent learning systems

“…There are several promising approaches to the learning path design: traditional solutions, where each learning material is finished with a question-and-answer session or knowledge evaluation test to the sequence of learning activities (Yang, Li, & Lau, 2010); competence tree transformation based design that allows students to choose different forms of the learning path (Margienė & Ramanauskaitė, 2021); reinforcement learning models for the learning path selection (Li, Xu, Zhang, & Chang, 2021). In each e-learning system, the type of learning path design can be adjusted based on the student's needs.…”