In the last decade many systems and models have been proposed aimed at making e-learning easier and more sound. The goal of this thesis is to develop a better e-NOTEBOOK system that improves collaboration and recommendation in group learning settings. To achieve this goal we have proposed an inquiry based collaborative learning web environment (Fig.7.1, Fig.7.2). The focus of proposed approaches is to support group learning by assisting learners in advancing their knowledge and skills through constructivist activities. Such environment must provide functions that promote learning activities and learning outcomes. This thesis focuses on promoting each learner’s contribution and learning outcomes, and therefore achieving efficient and effective group learning. The proposed e-NOTEBOOK system is based on the framework of activity theory.
144 Chapter 7: Conclusions and Future Works
(1) Inquiry based collaborative learning web environment: We have proposed a web based e-NOTEBOOK system. The system is oriented to support group learning and to provide a scaffold for learners as they learn to conduct collaborative, open-ended investigations on the web. The system provides a collaboration mechanism that allows for learner’s individual knowledge to be easily used and to easily trace the diffusion of the knowledge among learners. At the same time, it provides learners with dynamic recommendation of learning resources and recommendation of advanced learners, improves online information discovery and provides opportunities to socialize with other learners. By using these mechanisms, learners’ knowledge construction, knowledge production and knowledge acquisition can be facilitated. The system focuses on the participations of learners and learners’ knowledge constructing, acquisition and sharing in online. Notes and comments provided by learners are as products and contributions utilized to present learners’ knowledge, at the same time, they are as learning resources providing opportunities for constructing new and rich understanding. The learners expressed positive opinions regarding the contributions of our system.
Group learners
・Searching on web
・Note-taking, comment-taking
・Note-reading, comment-reading
Recommendation of advanced learners based on analysis of social interactions
•Interaction indicators
•Learners’ comparison
Recommendation of learning resources
•Content filtering based on keyword map
•Collaborative filtering based on learner’s relationship
Advanced learner group
Learning process Learning activity
Learning process Learning activity
Fig.7.1 Proposed main approaches and keywords
Chapter 7: Conclusions and Future Works 145
e-NOTEBOOK System
Inquiry Based Collaborative Learning Web Environment
1. Learning Environment
Knowledge construction
Sharing space Facilitation of knowledge acquisition
Facilitation of knowledge acquisition Knowledge
production
Inquiry cycle
Recommendation of Advanced Learners Recommendation of Learning Resources
Learners’
comparison profile
Knowledge production
& knowledge construction Estimate learners’
comparison
Web pages Notes &
Comments AbstractNOTE Remark
comment Web page Estimate interaction
indicators Learning processes
& learning activities
…
…
…
……
…… …
Estimate comparison
indicators Definition &
Calculation
Markov Chain Model
Natural language process &
data mining
User profile
| 遺産 | 1 |
| 安定 | 2 |
| 意匠 | 1 |
……
| 影響 | 3 |
| 回復 | 1 |
| 規模 | 会社 | 0.8698 |
| 規模 | 家屋 | 2.7783 |
| 規模 | 温室 | 1.3803 |
| 近代 | 漁獲 | 2.4387|
| 近代 | 激減 | 1.3330 |
……
| 地球 | 温暖 | 3.6650 |
| 燃料 | 石炭 | 1.9145 |
| 燃料 | 石油 | 2.0305|
| 燃料 | 生息 | 1.6378 |
| 燃料 | 太陽 | 4.3227 |
| 燃料 | 地熱 | 3.7207 |
lr4 lr7 lr1 lr5 lr8 lr12
1 2 3 4 5 6
keyword
relations Keyword map
Learning resources
Id: 1 Keyw ord map keyword
relations
| 洪水| 森林| 0.8228 |
……
| 騒音 | 1 |
……
Id: 2 Keyword map keyword
relations
|中国|汚染| 0.5524 |
……
|途上 | 2 |
……
Id: n Keyword map keyword
relations
|燃料| 地熱| 3.7607 |
……
| 騒音 | 1 |
……
……
Recommendation
Id: 3 Learni ng resource S(P3,LR3)=0.3121
å å
=
=ND= k
jk ik ND k
jk ik j i
M M
M M LR P S
1 1
) , max(
) , min(
) , (
Id: 2 Learni ng resource S(P3,LR2)=0.1342
Id: 6 Learni ng resource S(P3,LR6)=0.1109
Id: n Learning resource S(P3,LRn)=0.1311 Id: 9 Learni ng resource S(P3,LR9)=0.3733
……
Recommendation Recommendation u3
Estimated comparison Learner X Learner Rating
Learning Resource X Learner User 1
User 2
User
3 …. User
User n 1User 2User 3… User n User
1 User 2 User
3 …. User
Lr1 n Lr2 Lr3
… Lrm
User profile Group (Community)
lr4lr7lr1lr5lr8lr12
u3 123456
lr9 n
……
Learning process
Ｌｒ R
User preference
Predicted Ratings MXN Ratings predicting
User1 User2User3…. Usern Lr1 0.12270.0000.1138 ……0.1314 Lr2 0.009 0.21220.0117 ……0.1256 Lr3 0.21230.0000.1201 ……0.2232
…… …… …… …… …… …..
Lrm 0.12010.000 0.000……o.1101
Recommendation Id:23=0.0476;Id:47=0.5636;
Id:77=0.4321;Id34=0.0011……
Advanced learners
2. Recommendation of Advanced Learners 3. Recommendation of Learning Resources
(a)
(b)
Fig.7.2 Inquiry based collaborative learning web environment
(1) We proposed an inquiry based collaborative learning for knowledge construction, knowledge production and facilitation of knowledge acquisition; (2) We proposed a new approach for recommendation of advanced learners based on analysis of social interaction; (3) We proposed recommendation of learning resources takes into account both of the crucial components: the learning process and social interaction.
146 Chapter 7: Conclusions and Future Works
And knowledge construction should be facilitated by using the recommendation mechanisms. The proposed system allows the learner to construct knowledge and obtain experience through active participation in the learning environment. The learner should be engaged in solving meaningful problems through knowledge and product creation, and be obtained highly benefits. Learner learns from and with other group learners. Meanwhile, the system offers the personalized recommendations of learning resources and recommended advanced learner to collaborate with in group learning to facilitate knowledge acquisition.
(2) Recommendation of advanced learners: Participating in our system allows developing two aspects of collaborations: (a) peer interaction; and (b) expert to novice interaction. Knowledge collaboration makes progress through providing and receiving knowledge among learners. In our system, learners take on more roles. Besides being authors, presenters, they are peers and comment on the work of others. Each learner acts as both a knowledge producer and a knowledge receiver.
Construction of knowledge and skills result in an increased ownership of the learner in the learning process. Thus it is more group orientation and cooperation. We estimate interaction indicators (such as comprehension of web page, adequacy of remark, and agreement of comment) based on learners’
learning behaviors such as note taking, comment taking and etc. when group learners perform learning using in this system. Comparison indicators are defined as to describe a situation in which knowledge and understanding are socially distributed among group learners based on extracted interaction indicators. The integration of interaction indicators represents an important ingredient of this thesis. Moreover, our system evaluates contributions made by each learner using Markov Chain Model based on defined comparison indicators and recommends advanced learners to collaborate with. Since the learners may assume a role of experts and novices (depending on topics), the proposed system allows for productive collaboration among group learners. And via this mechanism, it achieves more active learning. In addition, more authenticity of the problems to be tackled can be achieved based on the recommendation from advanced learners.
(3) Recommendation of learning resources: With the rapid increasing of learning resources in e-learning, undoubtedly, recommender systems are becoming increasingly popular, owing to their
Chapter 7: Conclusions and Future Works 147
ability to reduce complexity and provide personalized environment for the learner. Current researches greatly benefit from the implementation of traditional recommendation technologies.
When dealing with personalized systems, we, as designers and researchers, need to reflect on how learners perceive and interact with the system and other learners. Our analysis shows that recommendation of learning resources is hybrid and it takes into account both of the crucial components: the learning process and social interaction. Our research on hybrid recommendation expanding the current research focus on two points:
3-(a) Content filtering based on keyword map to giving a complete user profiles, and adequate mechanisms for representing the information of the items and increasing both the relevance and complement of learning resources recommendation. Modeling the information about the learning process is closely connected to guidelines from educational, psychological, social and cognitive sciences. On the other hand, from educational aspect, this mechanism is mainly aids in inquiry reasoning based on transfer of learning in our system.
3-(b) Collaborative filtering based on learner’s relationship which considers the learners’
knowledge level to provide asymmetric interpersonal influence recommendation with implicit method. In this mechanism, we consider the advanced learners’ contributions to improve the accurate of learning resources recommendation thereby learners can select problems and learning resources more authenticity and raise questions they find worth considering. Moreover, from educational aspect, this mechanism provides a collaborative decision making in our system.
These approaches proposed by this dissertation that can overcome most of the problems faced by previous approaches, while achieving better prediction accuracy than conventional approaches.
These approaches allow for more flexibility and control on the recommendation process especially well suited for e-learning domain.
Through theoretical and experimental studies we have shown the effectiveness and applicability of the proposed approaches and the system that integrates them.