This paper introduces a hybridization approach of AI techniques and statistical tools to evaluate and adapt the e-learning systems including e-learners. Learner’s profile plays a crucial role in the evaluation process and the recommendations to improve the e-learning process. This work classifies the learners into specific categories based on the learner’s profiles; the learners’ classes named as regular, workers, casual, bad, and absent. The work extracted the statistical usage patterns that give a clear map describing the data and helping in constructing the e-learning system. The work tries to find the answers of the question how to return the bad students who are away back to be regular ones and find a method to evaluate the e-learners as well as to adapt the content and structure of the e-learning system. The work introduces the application of different fuzzy clustering techniques (FCM and KFCM) to find the learners profiles. Different phases of the work are presented. Analysis of the results and comparison: There is a match with a 78% with the real world behavior and the fuzzy clustering reflects the learners’ behavior perfectly. Comparison between FCM and KFCM proved that the KFCM is much better than FCM.
The design and implementation of web-based education (e-learning) systems have grown exponentially in the last years, spurred by the fact that neither students nor teachers are bound to a specific location and that this form of computer-based education is virtually independent of any specific hardware platforms (Romero, Gonza´lez, Ventura, del Jesus, & Herrera, 2009). These systems accumulate a great deal of information; which is very valuable in analyzing students’ behavior and assisting teachers in the detection of possible errors, shortcomings and improvements. However, due to the vast quantities of data these systems can generate daily, it is very difficult to manage manually, and authors demand tools which assist them in this task, preferably on a continuous basis. The use of data mining is a promising area in the achievement of this objective (Romero & Ventura, 2007). In the knowledge discovery in databases (KDD) process, the data mining step consists of the automatic extraction of implicit and interesting patterns from large data collections. A list of data mining techniques or tasks includes statistics, clustering, classification, outlier detection, association rule mining, sequential pattern mining, text mining, or subgroup discovery, among others (Klösgen & Zytkow, 2002). In recent years, researchers have begun to investigate various data mining methods in order to help teachers improve e-learning systems. A review can be seen in Romero and Ventura (2007); these methods allow the discovery of new knowledge based on students’ usage data. Subgroup discovery is a specific method for discovering descriptive rules (Klo¨ sgen, 1996 and Wrobel et al., 1997).
The proposed system for evaluating the e-learning systems and e-learners is shown in Fig. 1. The development of this system is our goal in this paper.
The rest of this paper is organized in the following way: Section 2 introduces survey on soft computing in e-learning. Section 3 describes the problems and goals of the work presented in this paper. Section 4 introduces the theoretical review for the fuzzy clustering techniques used. Section 5 introduces the data sets, including data preparation, data cleaning, data normalization, and features selection. Section 6 introduces the statistical analysis tools of web log files. Section 7 presents the experiments design and results analysis. Comparison between the different clustering techniques and the comparison with the real behavior are introduced in Section 8. Section 9 introduces the combination between the results of fuzzy clustering and the log file analyzer results to construct the suggestions and the recommendations for the e-learning system. Finally, the conclusions and further research are outlined in Section 10.
The work presented in this paper focuses on how to find a good methodology for the evaluation and adaption of the e-learning systems. The paper introduces the use of two techniques the first is the statistical tool for analyzing the log file and the second is the most important models in data mining; the clustering models; which were built using fuzzy clustering methods with the FCM and the KFCM. Both of FCM and KFCM clustering were able to find the clusters for the learners and the results were matched with the real marks of the students. Moreover, the KFCM results have high matching percentage with the real marks than the FCM. The application of the statistical tool as Logan pro software capable of discovering good statistical patterns as the traffic analysis during different time periods (monthly, weekly, daily), the directories and the type of files; that the learners were interested to access associated with the number of clients and the Bytes downloaded as well. The suggestion and the recommendations were constructed based on the clustering that represents the learner’s profiles and reflect their behavior during the teaching of the e-course, the discovered patterns obtained from the Logan pro tool, and the questioners obtained from the students themselves. Finally the paper proved that the ability of fuzzy clustering generally and KFCM was better, and the combination between the statistical tools and the AI or data mining methods provides better recommendation than one of them alone.
