DIFFERENT METHODS TO IDENTIFY STUDENTS PREFERENCES TO LEARNING STYLES AND LEARNING PATHS
Vilnius Gediminas Technical University (LITHUANIA)
About this paper:
Conference name: 12th annual International Conference of Education, Research and Innovation
Dates: 11-13 November, 2019
Location: Seville, Spain
Abstract:The aim of the paper is to analyse methods of identification of students’ personal preferences to learning styles and optimal learning paths. Students’ personal preferences have to be taken into account while creating optimal learning paths in order to achieve higher students’ motivation. Different dedicated psychological questionnaires for different learning styles models are presented in the paper. Several educational data mining (EDM) / learning analytics methods to identify and check students’ learning styles are also discussed. Additionally, exemplar models (neighbour method and case-based reasoning (CBR)) and Bayes networks (BN) are presented and discussed in more detail.
It is known that individuals use rules when new items are confusing and use exemplars when they are distinct. Initially, categorisation is based on rules. During the learning process, appropriate features for discriminating items is learned over time. Then, new items can be stored as exemplars and used to categorise less important items without discrepancies between rules. Exemplar models explain real life events that are problematic for modelling that uses sets of rules. New exemplar of the event is classified according to its similarity to the exemplars already stored.
Best known among examples of exemplar-based modelling are nearest neighbour method and CBR. The CBR method uses old experiences and adapts them for finding a solution to new problems.
In the paper, a systematic review of literature on modelling approaches combining CBR and BN is presented trying to identify current status of the development of the framework for Bayesian case-based reasoning. The literature review focuses on exemplar-based approaches, exploring possibilities of combining BN and CBR and seeking for niches for improvement of the overall BN-CBR approach.
In the paper, comparative analysis of existing CBR-BN models is also done. Attention is paid to feedback issues.
Finally, after discussion and weighting the pros and cons of application of combined BN-CBR approach as well as EDM methods for student’s learning style diagnosis and check, conclusions are made and future research trends are presented.
Keywords: Learning personalisation, learning styles, learning paths, psychological questionnaires, educational data mining, exemplar-based model, case-based reasoning, nearest neighbours, Bayes network.