Nor Bahiah Ahmad

A comparative analysis of mining techniques for automatic detection of student's learning style

This paper compares performance of several classifiers provided in WEKA such as Bayes, decision tree and classification rules in classifying students learning style. The students preferences and behavior while using e-learning system have been observed and analyzed and twenty attributes have been selected to map into Felder Silverman learning style model. There are four learning dimensions in Felder Silverman model and this research integrates the dimensions to map the students characteristics into sixteen learning styles. A 10-fold cross validation was used to evaluate the classifiers. Among parameters being observed in the performance of the classifiers are classification accuracy, Kappa statistics, training errors and time taken to build the model. The experiment showed that the tree classifiers have high accuracy with more than 91% accuracy. The sizes of the tree and the number of leaves among the tree classifier techniques have also been observed.

A Concise Fuzzy Rule Base to Reason Student Performance Based on Rough-Fuzzy Approach

A fuzzy inference system employing fuzzy if then rules able to model the qualitative aspects of human expertise and reasoning processes without employing precise quantitative analyses. This is due to the fact that the problem in acquiring knowledge from human experts is that much of the information is uncertain, inconsistent, vague and incomplete (Khoo and Zhai, 2001; Tsaganou et al., 2002; San Pedro and Burstein, 2003; Yang et al., 2005). The drawbacks of FIS are that a lot of trial and error effort need to be taken into account in order to define the best fitted membership functions (Taylan and Karagozoglu, 2009) and no standard methods exist for transforming human knowledge or experience into the rule base (Jang, 1993).

Developing Student Model Using Kohonen Network in Adaptive Hypermedia Learning System

This paper presents a study on method to develop student model by identifying the students’ characteristics in an adaptive hypermedia learning system. The study involves the use of student profiling techniques to identify the features that may be useful to help the researchers have a better understanding of the student in an adaptive learning environment. We propose a supervised Kohonen network with hexagonal lattice structure to classify the student into 3 categories: beginner, intermediate and advance to represent their knowledge level while using the learning system. An experiment is conducted to see the proposed Kohonen network’s performances compared to the other types of Kohonen networks in term of learning algorithm and map structure. 10-fold cross validation method is used to validate the network performances. Results from the experiment shows that the proposed Kohonen network produces an average percentage of accuracy, 81.3889% in classifying the simulated data and 51.6129% when applied to the real student data.

Meaningless to meaningful Web log data for generation of Web pre-caching decision rules using Rough Set

Web caching and pre-fetching are vital technologies that can increase the speed of Web loading processes. Since speed and memory are crucial aspects in enhancing the performance of mobile applications and websites, a better technique for Web loading process should be investigated. The weaknesses of the conventional Web caching policy include meaningless information and uncertainty of knowledge representation in Web logs data from the proxy cache to mobile-client. The organisation and learning task of the knowledge-processing for Web logs data require explicit representation to deal with uncertainties. This is due to the exponential growth of rules for finding a suitable knowledge representation from the proxy cache to the mobileclient. Consequently, Rough Set is chosen in this research to generate Web pre-caching decision rules to ensure the meaningless Web log data can be changed to meaningful information.

Hybrid NSGA-II of Three-Term Backpropagation network for multiclass classification problems

Hybridization has become one of the current focuses of new research areas of the evolutionary algorithms over the past few years. Hybridization offers better speed of convergence to the evolutionary approach and better accuracy of the final solutions. This paper presents a hybrid non-dominated sorting genetic algorithm-II (NSGA-II) to optimize Three-Term Backpropagation (TBP) network in terms of two objectives which are: accuracy and complexity of the network. Backpropagation algorithm (BP) is often used as a local search algorithm and when combined with NSGA-II, the performance of NSGA II is enhanced due to the improvement of the individuals in the population. The experimental results show that the proposed method is effective in multiclass classification problems. The results of the hybrid approach to the classification problems are compared with multiobjective genetic algorithm based TBP network (MOGATBP) and some methods found in the literature. Moreover, the results indicate that the proposed method is a potentially useful classifier for enhancing classification process ability.

Identifying Sign of Grit among Self-Paced MOOC Students Based on Clickstream Data

Previous research that investigate and predict student’s performance share various results concerning the factors that drive the students to retain and success. The factor ranging from intrinsic to extrinsic. However, few MOOC research attempt to scrutinize grit factor. Previous researches results are from survey. However, self-report survey is insufficient because the students may have different interpretation about the questions and measurement for themselves. Therefore, this is a case study that attempts to investigate the sign of grit among self-paced MOOC students using proposed features based on clickstream data. Also, the study observes whether the grit traits have relation with student’s performance. This study found that many students have high grade despite having low grit traits. However, statistical analysis shows that there is still significant correlation between grit traits and student’s grade. Other factor like unrecorded effort outside online learning may contribute to this result which can be investigated by future study. There is potential for better result by combining proposed method with collected survey.

Learning environment for software product lines online learning applications

The process of learning and teaching online learning has undergone many changes in line with technological developments. Education institutions have begun introducing new methods of learning this. However, it needs a huge amount of labor intensive to produce and maintain educational technologies due to its huge size (literacy, vocational education, school education, engineering and medical education) and huge variants (language, dialect). With the growing demand and at the same time would like to reduce the factor of cost, time and effort is long, then the need for an effective solution allowing rapid system development. A Software Product Line (SPL) approach is one of the best methods that can be used to develop an educational software family. The research contributes to the development of online learning SPL by studying issues in online learning SPL, identify appropriate pedagogical approach to online learning environment that is constructivism, and designing constructivist learning environment based on the theory of constructivism and design principles. In this paper, comparison between 10 VLE was tabulated against constructivist learning environment. The comparative evaluation is to discover commonalities and variabilities between online learning applications. The results should be useful as the process of domain analysis that is to document the commonalities and variabilities between SPL members. Hence, it can be used to enhance the feature model for online learning applications.