王志權

Abstract

An Adaptive Learning Guidance for Collaborative E-Learning: Mathematics Learning of Special Education Students as an Example

Zhi-Quan Wang

Yuh-Min Chen

Hui-Chuan Chu

Institute of Manufacturing Information and Systems

National Cheng Kung University

SUMMARY

This work develops an adaptive learning guidance for use in the benefit of students with special education needs a collaborative e-learning environment. When such students are learning, a model of the knowledge of a learning group is used to create a dynamically adaptive learning path. A genetic algorithm is used to group students. Then, an Apriori algorithm is used to generate the group knowledge model. The developed knowledge model that uses the developed algorithm to generate the adaptive learning path. Finally, a collaborative e-learning platform is generated to enable students to learn via the learning path. Experimental results indicate that groups that received adaptive learning guidance exhibited improved learning performance. The experimental results herein also revealed that students with moderate knowledge gained greater benefit than those with more knowledge.

Key words: Collaborative learning, E-learning, Adaptive learning, Heterogeneous grouping, Learning path

INTRODUCTION

In recent years, countries around the world have begun to focus on the education of disadvantaged students, actively promoting policies to help disadvantaged students to catch up with other students. More advanced countries have longer historical periods of the development of special education, and special education students in such countries receive more attention. Special education students differ in physical and mental developmental and learning-related characteristics, as well as other respects, from average students. Because learning for such students is harder than for general students, learning outcomes are relatively poor, and those students are more difficult to teach using traditional methods. Teachers must pay correspondingly more attention to them.

Research (Rovai, 2000) has demonstrated that e-learning differs from traditional classroom learning, and can help special education students to learn more effectively. However, most e-learning systems lack appropriate learning guide mechanisms, so learners are likely to experience cognitive overload and disorientation, reducing the effectiveness of learning. Slavin (1985) regarded cooperative learning as the goal of structured systematic teaching strategies. If teachers do not understand their students well, and they fail to teach in a manner that takes into account the heterogeneity of their students, then learning suffers.

E-Learning

E-Learning, formerly known as distance learning, emphasizes the use of electronic media to transmit teaching content, two-way interactive teaching, and the creation of a digital learning environment.

Cooperative learning

Cooperative learning emphasizes heterogeneous groupings and group cooperation. Team members work together to accomplish a common goal. Some scholars (Dugan, Kamps, Leonard, Watkins, Rheinberger, ＆ Stackhaus, 1995) have utilized cooperative learning to help autistic children in two fourth-grade classes, and they found an increase in the weekly quiz scores of individual children with autism during the period of intervention, improved participation and focus, relative to a reference period, and increased social interaction with peers.

Computer-supported Collaborative Learning

Research over the last decade has confirmed that the network can increases the number of opportunities that students can take to participate actively in cooperative learning. Scifres ＆ Behara (1998) asserted that allowing remote network team members to interact continuously is important to a network of cooperative learning.

This work develops the use of genetic algorithms to optimize a cooperative learning team. In this research, knowledge structures are constructed by Apriori algorithm.

To verify that the proposed adaptive e-learning guidance system improves the cooperative learning outcomes of students, purposive sampling is used. Experimental samples are obtained from 20 high-school special education students. The experimental group exhibited better mathematical performance and collaborative performance than the control group.

MATERIALS AND METHODS

Heterogeneous grouping

In this work, we regard the grouping problem as an optimizing team problem. The optimizing team problem is one kind of best combination problem. We use the genetic algorithm which can be the best kind of lieutenant colonel algorithm to solve this optimizing team problem.

Learning path planning

This study analyzes the differences between standard knowledge model and learning group knowledge model. According to Ling-Hsiu Chen (2011) proposed misconceptions diagnostic methods, we added the condition threshold filtering method to increase the accuracy of the misconceptions diagnosis method. According to the results of the difference analysis in knowledge model, we use the proposed algorithm to select learning materials according to the misconceptions. Finally, we generate an adaptive learning path for students to learn.

RESULTS AND DISCUSSION

Heterogeneous grouping

According to students learning style score and thinking style score, we use this two scores to generate heterogeneous group. Genetic algorithms are used herein to perform the heterogeneous grouping.

Figure 5.9 Heterogeneous grouping results

Figure 5.9 presents the sum of the deviation from the mean (SDFM) as calculated using genetic algorithm. A minimum SDFM is obtained, representing the best grouping, as shown in Table 5.4.

Table 5.4 Heterogeneous grouping results

Gruop Members

1 16,8,18,14

2 11,4,5,7

3 2,3,17,6

4 13,10,15,19

5 1,9,12,20

CONCLUSION

E-learning represents a new trend. Cooperative learning not only help students to solve problems but also improves their ability to interact with others. This research proposes an adaptive digital technology cooperative learning model for students who require special teaching. The main contributions of this research are adaptive cooperative e-learning guidance, a method for heterogeneously grouping students and a method for planning learning paths. Experimental results revealed that the students who were taught using this system performed better than the control group in mathematics. This work considered only students with mild disabilities, whose bodily functions do not differ from those of the general population, and who can learn to operate a computer. However, many special education students are limited in their ability to use computer hardware and software. Students with special may need hardware with special configurations, and improved user interfaces to support them in digital cooperative learning.