Paper Title : A Hybrid Classification Method for Disengagement Detection in Online Learning
Author(s): P. V. Praveen Sundar, A.V. Senthil Kumar
Published in: International Journal of Education and Information Studies (IJEIS)
Volume/Issue: Vol. 05, Issue 01 ( 2015), ||V1|| PP 67-74
ISSN : 2277- 3169
Abstract : The main aim of the online learning system is to meet the requirements of the learners and to make efficient for learners where the aspects and complexity are taken into consideration. The learner’s motivational states are undertaken by many attempts, mainly by using design. Motivations are started by using analysis of log file. Firstly, the disengaged learners are identified moderately, and then visualize the disengaged learners which includes evaluation of many motivational characteristic for learning. For improvement in learning, data mining and machine learning methods will provide us meaningful data and valuable information. The performances of Bayesian classifiers endure in the field where it involves correlated features. Naïve Bayesian classification with PSO method is already implemented in many fields, the main problem in PSO is its tendency of trapping into local optima. To overcome this problem, this research presents the hybrid algorithm by combining fast PSO and Naïve Bayesian classifier for classification to aid in the prediction of disengagement. According to the characteristics of the data, our proposed method improves the classification accuracy and avoids the loss of information. This study results showed that the method was feasible and effective.
Keywords-- Online learning, Log File Analysis, Disengagement Detection, Bayesian classifiers, Particle Swarm Optimization, Quasi Framework.
Download Full Text
To Cite this paper
P.V. Praveen Sundar, A.V. Senthil Kumar,"A Hybrid Classification Method for Disengagement Detection in Online Learning", International Journal of Education and Information Studies (IJEIS), Volume 5, 2015 PP: 67-74.
Author(s): P. V. Praveen Sundar, A.V. Senthil Kumar
Published in: International Journal of Education and Information Studies (IJEIS)
Volume/Issue: Vol. 05, Issue 01 ( 2015), ||V1|| PP 67-74
ISSN : 2277- 3169
Abstract : The main aim of the online learning system is to meet the requirements of the learners and to make efficient for learners where the aspects and complexity are taken into consideration. The learner’s motivational states are undertaken by many attempts, mainly by using design. Motivations are started by using analysis of log file. Firstly, the disengaged learners are identified moderately, and then visualize the disengaged learners which includes evaluation of many motivational characteristic for learning. For improvement in learning, data mining and machine learning methods will provide us meaningful data and valuable information. The performances of Bayesian classifiers endure in the field where it involves correlated features. Naïve Bayesian classification with PSO method is already implemented in many fields, the main problem in PSO is its tendency of trapping into local optima. To overcome this problem, this research presents the hybrid algorithm by combining fast PSO and Naïve Bayesian classifier for classification to aid in the prediction of disengagement. According to the characteristics of the data, our proposed method improves the classification accuracy and avoids the loss of information. This study results showed that the method was feasible and effective.
Keywords-- Online learning, Log File Analysis, Disengagement Detection, Bayesian classifiers, Particle Swarm Optimization, Quasi Framework.
Download Full Text
To Cite this paper
P.V. Praveen Sundar, A.V. Senthil Kumar,"A Hybrid Classification Method for Disengagement Detection in Online Learning", International Journal of Education and Information Studies (IJEIS), Volume 5, 2015 PP: 67-74.
No comments:
Post a Comment