Classification Model of English Course e-learning System. for Slow Learners

Transcription

1 Classification Model of English Course e-learning System for Slow Learners Thakaa Z. Mohammad, Abeer M.Mahmoud, El-Sayed M. El-Horbart Mohamed I.Roushdy and Abdel-Badeeh M. Salem Department of Computer Science, Faculty of Computer and Information Sciences, Ain Shams University,Cairo,Egypt ABSTRACT: Data mining methodology can analyze relevant information results and produce different perspectives to understand more about the students activities. When designing an educational environment, applying data mining techniques discovers useful information that can be used in formative evaluation to assist educators establish a pedagogical basis for important decisions. This study presents a proposed model based on classification approach to find an enhanced evaluation method for slow learner students in the English course.this model can determine the relations between academic achievement of students and followed behaviors in the website of e- learning English course. In addition to that, the results of this study provide the teachers with new enhanced methods of evaluation since the slow learners might not have high academic achievements. Key-Words:- E-Learning, Knowledge Discovery in Databases, Data Mining, Web Mining, Slow Learners, English e-learning course. 1 Introduction E-learning is a technological educational method aims to develop the education by providing facilitated learning methods and approaches based on information, communication and multimedia technologies. During the last five years, a huge of accumulated information is collected from e- Learning courses [1,2]. Furthermore, applying data mining techniques on educational databases offers helpful and constructive recommendations to the academic planners in education institutes to enhance their decision making process, to improve students academic performance and trim down failure rate, to better understand students behavior, to assist instructors, to improve teaching and many other benefits [3,7,10]. On the other side, classifying those databases and finding out suitable methods to retrieve information from them created the needs for new approaches [6,8,9]. Classification is a process of grouping physical or abstract objects into classes of similar objects. It is a supervised model where it predicts class labels for unseen data. When using a classification for educational system, it allows characterizing the properties of a group of user profiles, similar pages or learning sessions. Recently, the classification is one of the most useful tasks in e- learning since many objectives can be achieved in e- learning environments by utilizing the classification task such as; (a) classifying students into groups according to the similarity in their reactions and characteristics as an educational strategy, (b) identifying students whom need more motivation as a kind of minimizing the rates of students dropping out, and (c) predicting the levels of students intelligence for giving them special courses, and many other objectives [3,15]. This paper discusses the application of data mining approach to educational systems, particularly, classification techniques for e-learning. This study aims to find the relation between the students behaviors in the websites and their academic achievements. Therefore, a data mining model ISBN:

2 based on classification methodology has been built for this purpose. However, we can define the relations between students behaviors in the e- learning website and their academic achievements. The rest of this paper is organized as follow: Related work is discussed in section 2 and our problem domain is described in section 3. Section 4 presents our proposed classification model for tracking students behaviors. In section 5, the paper concludes with a short summary and with an outlook on the next steps to be done in the proposed development system. 2 Related Work Many studies have been held in order to improve the implementation of data mining means and algorithms through e-learning systems [8,9,10,11]. Each study focuses on a special data mining technique as an application for e-learning. In what follows, a brief discussion is presented for some related work achieved in this respect. AlAjmi et al. [16] held a study about Using Instructive Data Mining Methods to Revise the Impact of Virtual Classroom in E-Learning. In their study, they have focused on the use of educational and instructive data mining approaches in finding out the impact of virtual classroom in e-learning which is considered as a kind of virtual learning environment. Also, they followed using a data mining techniques to observe the records of students behaviors in the virtual classrooms environment. Actually, they have classified the performance indicators of students as variables of the data mining algorithm. After that, they calculate the weights of those variables (for each student) based on exams scores and then, classify the students to conclude their study. Chellatamilan and Suresh [17] utilized the data mining tools in recording and classifying the behaviors of e-learning students in order to find the impacts of the e-learning systems on the students. Chen et al. [12] apply decision tree (C5.0 algorithm) and data cube technology from web log portfolios for managing classroom processes. The induction analysis discovers potential student groups that have similar characteristics and reaction to a particular pedagogical strategy. Minaei- Bidgoli and Punch [13] classify students based on features extracted from the logged data in order to predict their final grades. They use genetic algorithms to optimize a combination of multiple classifiers by weighing feature vectors. Tang et al. [5] use data clustering for web learning to promote group-based collaborative learning and to provide incremental learner diagnosis. They find clusters of students with similar learning characteristics based on the sequence and the contents of the pages they visited. Hamalainen et al. [4] introduce a hybrid model, which combines both data mining and machine learning techniques in constructing a Bayesian network to describe the student s learning process. ISBN:

3 First Semester Unit 1 al z al z al z al z al z Unit 2 al z al z al z al z al z English E-Learning Course Unit 6 Second Semester Unit 7 al z al z al z al z al z First Semester al z al z al z al z al z Unit 8 al z al z al z al z al z Unit 12 al z al z al z al z al z Second Semester English Course Figure 1 English Course in the Department of Slow Learners 3 Problem Domain By focusing on slow learners students, there are many courses that available for those students who suffer from some problems that make them learn in a slow manner comparing with their equals in normal classes. Through this study, a new English e-learning course for slow learners will be designed. Actually, this course will be designed to contain twelve different units divided into two semesters where unit- 1 to unit-6 forms the first semester and unit-7 to unit- 12 forms the second semester. Each unit will have five parts; reading, listening, grammar, speaking, and writing. Each part will contain two levels; material level that explains the idea of the part, and the quiz level that assesses the understanding level of the student for the required skills. A total score is calculated for each unit in the semester, and also a total score is calculated for each semester. Finally, the two semesters scores are used to calculate the average score of English e-learning course for each student as it appears in Fig 1. The English course is just an example about the courses that can be offered for slow learners for different levels. It appears that the academic achievement of each student will be noticed by the teacher based on the total score of the unit, semester, and the course. Containing five parts in each unit (Reading, Listening, Grammar, Speaking, and Writing) makes several choices for each student to follow different patterns in selecting the pattern that they prefer in the e-learning course. This e-learning course will not force a specific order of parts to follow by the students of slow learners in English e- learning classes. Some students start with the Reading part, other students start with the Listening part, and so on. ISBN:

4 Figure 2 The proposed classification methodology in tracking students behaviors Pattern 1: Reading Listening Grammar Speaking Writing Pattern 2: Reading Listening Grammar Writing Speaking Pattern 3: Reading Listening Speaking Grammar Writing Pattern 120: Writing Speaking Grammar Listening Reading Figure 3 Patterns of students behaviors in the e-learning environment These behaviors form different patterns that followed by the students in the same class. Actually, following the behaviors of the students within the course in order to compare those behaviors with their academic achievement cannot be done simply. 4 The Proposed classification model for tracking students behaviors. The solution of this problem can be achieved by the use of classification approach as shown in Fig 2. According to the figure, two types of data are required to find the relation between the patterns and the academic achievement of the students; data of navigating students behaviors and data of recording students academic levels. (a) The first kind of data will be classified into patterns since each pattern will have an ID as it is shown in Fig 3. Those patterns can be obtained simply by tracking students behaviors in the e- learning environment. The number of these patterns will be equal to 5*4*3*2*1 (5!) which equal 120 based on the probability rule; the tree in Fig 4 can explain this relation in more details. ISBN:

5 Student Main Screen z z z z z Listening Grammar Speaking Writing Reading Listening Speaking Writing Reading Listening Grammar Speaking Patterns Formation in each unit Figure 4 Patterns Formation in Each Unit Obtaining Relations: Pattern x the Highest Academic Achievement of Students Pattern y the Lowest Academic Achievement of Students Pattern z Average Score between m and n Figure 5 Final forms of obtained relations (b)the second type of data can be collected from the records in the databases for the involved students. Each student has a file that contains the academic scores in each part and each unit in addition to the records of accessing and exiting date and time. This database is used by the teachers to keep track with their students achievement. The collected data can be linked together in order to find the relations between patterns of students behaviors and their academic achievement by using classification techniques of data mining. The collected data will be compared in a table form, where each pattern is linked with the total achievement of the students that follows this pattern. (c) The final step which is in obtaining the relations between the linked data can be in statements form as it appears in Fig 5. However, the English e-learning course will be designed by HTML programming language in order to make it available on a website. Some GUI functions and methods will be used in order to design a user friendly interface. The HTML programming ISBN:

6 provides each page with a unique link and so it is easy to track students. Then, the obtained data from web mining techniques can be classified by simple relationships based on the previously defined patterns. 5 Conclusions This study presents a proposed model based on classification approach to find an enhanced evaluation method for slow learners students in the English course. This model can determine the relations between academic achievement of students and followed behaviors in the website of e-learning English course. In addition, the results of this study provide the teachers with a new enhanced methods of evaluation since the slow learners might not having high academic achievements; therefore, evaluating students with new methods can be more effective than traditional methods. References [1] E. García, C. Romero, S. Ventura, and C. Castro, Using Rules Discovery for the Continuous Improvement of e-learning Courses, Lecture Notes in Computer Science, vol. 4224, pp , [2] M. E. Zorrilla, E. Menasalvas, D. Marin, E. Mora, and J.Segovia, Web usage mining project for improving web-based learning sites, In Web Mining Workshop, pp. 1-22, [3] Kumar, V. and Chadha, A. An Empirical Study of the Applications of Data Mining Techniques in Higher Education, Int.J. of Advanced Computer Science and Applications, vol. 2, no. 3, pp , [4] Hamalainen, W., Suhonen, J., Sutinen, E., & Toivonen, H. Data mining in personalizing distance education courses. Int. conf. on open learning and distance education, pp. 1-1, [5] Tang, T., & McCalla, G. Student modeling for a web-based learning environment: A data mining approach. Proc. Of 18th conf. on artificial intelligence, USA, pp , [6] B. M. Ramageri, Data mining techniques and applications, J. of CS and Engineering, vol. 1, no. 4, pp , [7] J. Han and M. Kamber, Data Mining: Concepts and Techniques, 2 nd ED, Morgan Kaufmann Publishers, [8] P. Berkhin, Survey of Clustering Data Mining Techniques, 2002, retrieved 24th Sep, 2012 < [9] D. Keim and M. Ward, Visual Data Mining Techniques, 2002, retrieved 24th Sep, 2012 < 9/6984/ > [10] C. Romero and S. Ventura, Educational data mining: A survey from 1995 to 2005, int.j.of Expert Systems with Applications, vol. 33, pp , [11] Zorrilla, M. E., Menasalvas, E., Marin, D., Mora, E., & Segovia, J., Web usage mining project for improving web-based learning sites, In Web mining workshop, Cataluna, pp , [12] Chen, G., Liu, C., Ou, K., & Liu, B., Discovering decision knowledge from web log portfolio for managing classroom processes by applying decision tree and data cube technology, J. of E-Computing Research, vol.23, no.3, pp , [13] Minaei-Bidgoli, B., & Punch, W, Using genetic algorithms for data mining optimization in an educational web-based system. In GECCO, pp , [14] Nilakant, K., & Mitrovic, A. Application of data mining in constraint-based intelligent tutoring systems. In Proceedings of the artificial intelligence in education, AIED, pp , [15] Li, J., & Zaı ane, O., Combining usage, content, and structure data to improve web site recommendation. In Int. conf. on ecommerce and web technologies, pp , [16] M. AlAjmi, S. Khan, and A. Zamani, Using Instructive Data Mining Methods to Revise the Impact of Virtual Classroom in E-Learning, Int. J. of Advanced Science and Technology, vol. 45, pp , [17] T. Chellatamilan and R. Suresh, An e- Learning Recommendation System using Association Rule Mining Technique, European J.of Scientific Research, vol. 64, no. 2, pp , ISBN: