Data Mining of Digital Footprint during Assessment Grading for Intelligent Decision Making during Learning Process

Abstract

The paper describes the problems of generation and modeling of learning assessments, their grading, and analyzing the process of their completion by learners using data mining methods. The authors propose to enhance the digital footprint of learner’s interactions with e-assessment systems with the information allowing determining demonstrated level competence development and the causes of learners’ mistakes and hypo. The learner’s level of competence development can be evaluated using expert estimates and the Certainty Factor. Generating assessments and determining the causes of learner’s mistakes can be done using ontological models of subject domains, built on the comprehension level of Bloom’s taxonomy. The paper explores the properties of such models. Examples of applying the proposed approaches to intelligent decision making during the learning process of different subject domains are shown. This leads to formulating the requirements for subject domain models for intelligent tutoring systems