%0 Conference Paper
%A Bamatraf, S.
%A Aboalsamh, H.
%A Hussain, M.
%A Mathkour, H.
%A Qazi, E.-U.-H.
%A Malik, A.
%A Amin, H.
%D 2014
%F scholars:4302
%I Institute of Electrical and Electronics Engineers Inc.
%K Artificial intelligence; Brain mapping; Education; Electroencephalography; Principal component analysis; Support vector machines, Learning; PCA; Short term memory; SVM; Topomaps, Biomedical signal processing
%P 187-192
%R 10.1109/AIMS.2014.59
%T Studying the Effects of 2D and 3D Educational Contents on Memory Recall Using EEG Signals, PCA and Statistical Features
%U https://khub.utp.edu.my/scholars/4302/
%X Learning and memory are two related mental processes. EEG is a brain mapping technique, which can record brain states directly and can be used to assess learning and memory recall. In this paper, we will assess the effects of 2D and 3D educational contents on learning and memory recall by analyzing the brain states during recall tasks using EEG signals. 34 subjects learn same 2D and 3D educational contents and after 30 minutes, they are asked multiple-choice questions (MCQs) related to the learned contents. Though the answers of MCQs can be used to assess the effects of 2D and 3D educational contents on learning and memory recall, the correct answer of an MCQ can be based on just a guess. We studied direct brain states by analyzing EEG signals for this purpose and modeled it as a classification problem. While answering an MCQ, EEG signal is recorded, which is then converted into topomaps. The number of topomaps corresponding to one MCQ is excessively high and there is a large number of redundant topomaps, Principle Component Analysis (PCA) is used to reduce this number. Finally, statistical features are extracted from these topomaps and passed to Support Vector Machine (SVM) to predict brain states corresponding to correct/incorrect answers. The results of the study showed that 3D content gave 81.6 classification accuracy compared with 76.1 given by 2D content. It indicates that 3D educational content is more effective than 2D educational content. The proposed system can used to predict the memory recall level of a subject, which can help to select the educational content and future carrier for a subject. © 2014 IEEE.
%Z cited By 2; Conference of 2nd IEEE International Conference on Artificial Intelligence, Modelling, and Simulation, AIMS 2014 ; Conference Date: 18 November 2014 Through 20 November 2014; Conference Code:112206