ADAPTIVE MODEL TO SUPPORT DECISION-MAKING IN LANGUAGE E-LEARNING

The paper deals with the area of e-learning education, primarily in adaptive learning of the English language. The research consists in proposing an expert system for decision-making support in defining an ideal learning model, individualised for each student. Students entering an e-learning course have both different input knowledge and preferred learning styles. Thus the system is designed in order to enable the student to pass automatically through the e-learning course while the condition of adapting the content to the student's learning style is met. Adaptation and automated passing are achieved by dividing the process of studying 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 M1 - M4 is solved by the decision-making system operating with vague information. Process (M1) 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. The decision-making process is also influenced by information of a non defined character (fuzzy), the system then chooses a solution using IF-THEN rules, which are provided from a knowledge base. The rules utilise expert knowledge on language learning/teaching provided by the course teacher/tutor. Correct setting of the rules, primarily in process (M1), is crucial for further decisions made in processes (M2-M4). Therefore, the adaptive system for decision-making support will enable automated creation of study variants which are suited to each individual students' needs, which current learning management systems do not enable.