Language e-Learning Based on Adaptive Decision-Making System

摘要

A model of adaptive e-learning system which is in the focus of this paper is created for significant optimisation of language learning with the primary focus on the English language, which is based on using Learning management systems (LMS).In comparison with traditional "static"/"rigid" models of e-courses, which actually offer a uniform, content to all students, the proposed model offers its users/students adaptation in the area of their knowledge level and learning style (their combination respectively). In addition, the model is supported by automated decision-making processes, which are the most significant .tool for optimising language learning. Automation of such decision-making processes is required in several areas where the decision-making processes are entered by input Information in order to be appropriately processed to obtain desired outputs for ideal progress in further steps of student's learning process. Such input information provided by the student carries a certain extent of uncertainty, thus it is necessary to base the decision-making processes on IF-THEN rules supported by a fuzzy-modelling tool. Adaptation and automated progress are achieved by dividing the progress into four basic processes - (M1) acquisition, completion and evaluation of the input information about the student; (M2) definition of language learning objectives and admissible solutions for their achieving; (M3) modelling the progress of individual proposed solutions; (M4) approving one of the proposed solution as the best variant to achieve the learning objectives. Each individual process is dealt with by the expert system individually, when the system evaluates the input information of the given process using a knowledge base, which is individually created for every step. The knowledge bases are defined by an expert/tutor. Process (Ml) contains two basic subprocesses dealing with acquiring information about the student: their learning style based on a questionnaire inquiry and their language level based on a didactic test. Correct setting of the expert system in process (Ml) enables the system to proceed to processes (M2), M3 and (M4). Information evaluated by process (M1) enters process (M2). This process is a step when the system, based on the defined objectives of the language learning, selects suitable learning objects which will be presented to the student in further steps. Such objects must be appropriately matched with the given learning styles and purposefully lead to the defined objectives. Process (M3) then models the proposed objects into a sequence in which they will be presented, their time requirements as well as evaluation of the overall study progress both from the point of view of didactics and time. The two processes are very closely related and their successful evaluation by the system leads to student's successful passing the course, which is represented by step (M4). Therefore, the adaptive system for decision-making support will enable automated creation of study variants which are suited to each individual student's needs, which current learning management systems do not enable.