Constructing an Adaptive Mobile Learning System for the Support of Personalized Learning and Device Adaptation

(1)

Procedia - Social and Behavioral Sciences 64 ( 2012 ) 332 – 341

INTERNATIONAL EDUCATIONAL TECHNOLOGY CONFERENCE

IETC2012

Constructing an Adaptive Mobile Learning System for the

Support of Personalized Learning and Device Adaptation

Ho-Chuan Huang

a

, Tsui-Ying Wang

b,*

, Fu-Ming Hsieh

a

a_{Department of Information Management, National Kaohsiung University of Applied Sciences, Kaohsiung 807, Taiwan, ROC} b_{Department of Occupational Therapy, National Cheng Kung University, Tainan 710, Taiwan, ROC}

Abstract

This paper presents an adaptive mobile learning system (AMLS) that provides learners with adaptive content according to their knowledge levels, learning styles, and heterogeneous learning devices. The aim of the proposed work is to provide learners with an adaptive learning environment according to learner's individual capability and the learning device used. The proposed system exploits Bayesian networks and content adaptation technologies to support both learner adaptation and device adaptation, which allows each learner to construct a personalized and adaptive learning environment.

Keywords: Adaptive learning; Content adaptation; Bayesian networks; Learning styles; Mobile learning.

1. Introduction

Mobile learning has received a lot of attention in education with the emergence of an increasing number of new types of mobile devices, such as notebooks, personal digital assistants (PDAs), and smart phones. Learners often wish to use various types of learning devices to access the same content without sacrificing usability and accessibility. However, most of the content in Web-based educational systems is

*_{Corresponding author. Tel.: +886-6-275-7575 ext. 5903; fax: +886-6-237-6604.} E-mail address: michwang@mail.ncku.edu.tw.

(2)

typically designed and optimized for desktop computers, which make it unsuitable for use with other types of learning devices to view the content. Even though some of the Web-based educational systems support mobile learning, they may offer content only on specific types of devices. Because content adaptation is one of the technologies that support adaptive versions of content for heterogeneous devices, there is an increasing demand for content adaptation for a Web-based learning environment. In addition to the discrepancies of learning devices, learners may have different abilities, preferences, motivations, and knowledge. Some Web-based systems are devoted to develop the techniques of content adaptation for the problem of device heterogeneities (Laakko & Hiltunen, 2005). However, learner profiles (e.g., learning styles and knowledge levels) should be considered for the design of personalized learning assistance. From a learner's point of view, adaptation results not only should satisfy individual demands (i.e., learner adaptation) but should also solve the problem of device discrepancies (i.e., device adaptation). Therefore, it is both important and challenging to adapt content so that it satisfies individual demands and fits the requirements of learning devices in a mobile learning environment.

In summary, learner adaption and device adaption are considered as two important factors to facilitate mobile learning environments for learners with various abilities and learning styles. This paper presents an adaptive mobile learning system (AMLS) that exploits both learner adaptation and device adaptation to construct a personalized learning environment according to the individual characteristics and abilities. The adaption technology used in AMLS is not new but this study proposed a novel approach which combines both learner adaption and device adaptation for the support of personalized mobile learning environment. Learner adaptation is defined as matching content to the abilities and preferences of individual learners. Device adaptation is defined as automatically adapting content to the capacities of heterogeneous learning devices. In this paper, learner adaptation refers to individual differences in both knowledge levels and learning styles, and device adaptation refers to the heterogeneities of learning device specifications.

2. Related Works

Different AES have been developed for various purposes of education. Adaptive Hypermedia Architecture (AHA) is a Web-based adaptive hypermedia system, which can support on-line courses with different adaptive features, such as conditional explanations and links (De Bra et al., 2003). The learner model in Adaptive Hypermedia Architecture (AHA) is based on concept knowledge obtained and evaluated by Web-based courses and testing (De Bra, Aroyo, & Cristea, 2004). Karampiperis and Sampson (2005) proposed an adaptive educational hypermedia system, which supports adaptive learning resource sequencing based on a decision model that chooses an adaptive learning resources by evaluating learner's abilities. Henze and Nejdl (2004) introduced a logical characterization for the definition of adaptive educational hypermedia systems (AEHS) as a quadruple (DOCS, UM, OBS, AC): DOCS (Document Space) describes documents and knowledge topics; UM (User Model) stores, describes, and infers individual user's information, knowledge, preferences; OBS (Observations) observes individual user's knowledge state and interactions with the system for updating UM; AC (Adaptation Component) contains rules for the describing the adaptive functionality of the system.

ANDES used BN (Bayesian Network) technologies to model learners' knowledge in Physics (Gertner & VanLehn, 2000). If a BN model diagnoses a learner who did not understand a knowledge concept, the learning assistance for that concept would automatically appear on the screen to help the learner. BITS, a Web-based Bayesian intelligent tutoring system, uses BN to model problem domains in programming languages and creates adaptive learning sequences for learners according to their knowledge levels (Butz, Hua, & Maguire, 2006). All of the above proposed systems have used BNs or probability computing

(3)

technologies to assist learners in mastering content, but none of them has considered both learner adaptation and device adaptation.

Content adaptation solves the problems of adapting content to heterogeneous device capabilities and supporting individual user preferences (Canali, Cardellini, Colajanni, Lancellotti, & Yu, 2003). Odyssey Systems (Noble, 2000) used static adaptation technologies to create pre-adapted content versions for specific learning devices. The advantage of static adaptation approach is that content transformation causes no delay in content delivery. One serious problem of this approach, however, is that it requires a new version of the content new type of learning device accessing the system. The more heterogeneous learning devices that the static adaptation approach must support, the more expensive and time-consuming it becomes to create different versions of the same content. On the other hand, dynamic adaptation dynamically generates the desired content based on the specifications of heterogeneous devices. Multiple pre-adapted versions of the content need not be created or stored, but a transcoding mechanism is required for dynamic content transformation (Liang et al., 2006). The major advantages of the dynamic adaptation approach are that it offers great flexibility in the support of heterogeneous learning devices and avoids the inconsistent content almost certain to appear in multiple versions made for different devices. Kim and Lee (2006) proposed a content adaptation architecture that integrates Composite Capabilities/Preference Profiles (CC/PP) files and annotation mechanisms to dynamically construct Web pages by annotating and reconstructing the structure of Web elements for mobile devices. They also developed a navigation map to decide which elements should be contained in the adapted content.

3. System Overview

3.1. System architecture

The architecture of the AMLS consists of six modules (see Fig. 1):

• The user interface module is a graphic interaction interface between AMLS and the device on the client side.

• The context detection module is responsible for detecting context information, which includes the monitoring of a learner's progress and behavior and the detection of learning devices.

• The learner profile module manages information related to an individual learner's demographic details, such as learning preferences and learning styles. It also manages a learner's states obtained from both the context detection module and the learning diagnosis module.

• The learning diagnosis module is composed of a knowledge-diagnosis mechanism and a style-diagnosis mechanism. The knowledge-style-diagnosis mechanism evaluates learner's knowledge levels by comparing learner's knowledge in the learner model against the expert knowledge in the knowledge model. The style-diagnosis mechanism identifies individual learning styles based on the information obtained from the learner profile module.

• The expert knowledge module stores the expert knowledge in the knowledge database to support knowledge diagnosis. This module also contains the learning materials and relevant pedagogical strategies to assist individual learners.

• The content adaptation module is responsible for presenting the adapted learning content. The adaptation process includes learner adaptation and device adaptation.

(4)

Table 2. Questionnaire result for participants (n=25)

No Statement Strongly

Agree Agree Neutral Disagree Strongly Disagree

Mean (SD) 1 It is easy to access the learning content

on my mobile device. 11 (44%) 14 (56%) 0 (0%) 0 (0%) 0 (0%) _(0.51)4.44 2 I am satisfied with the arrangement of

learning content on my mobile devices. 6 (24%) 16 (64%) 3 (12%) 0 (0%) 0 (0%)

4.12 (0.60) 3 The learning content images fit well on

my mobile screen. 6 (24%) 14 (56%) 5 (20%) 0 (0%) 0 (0%)

4.04 (0.68) 4 I can quickly locate the learning content

using the navigation function. 8 (32%) 13 (52%) 4 (16%) 0 (0%) 0 (0%)

4.16 (0.69) 5 Overall, I am satisfied with my learning

experience when using mobile devices. 6 (24%) 16 (64%) 3 (12%) 0 (0%) 0 (0%)

4.12 (0.60) 5. Conclusions

The support of heterogeneous mobile devices is important for increasing learning convenience and efficiency in a mobile learning environment. By identifying individual device capabilities, content adaptation provides a solution to the heterogeneity of devices for learners. In an adaptive educational system, content adaptation offers appropriate learning content suited both to the device's specifications and to the learner's abilities. Therefore, learning diagnosis is an important procedure for identifying the preferences and knowledge levels. This study proposes an adaptive mobile learning system that uses adaptation to both the learner and the learning device to create a personalized and adaptive learning environment suited to the learners' abilities and the device's specifications. A learning diagnosis mechanism was constructed to diagnose each learner's knowledge levels and identify each learner's learning styles. In addition, content adaptation technologies were also used to automatically adjust content to match the specifications of learning devices. Further research is encouraged to improve the inference capability when managing a learning context and arranging content in heterogeneous learning devices.

Acknowledgements

This work was supported in part by National Science Council (NSC), Taiwan, under the Grants NSC100-2511-S-151-001-MY2.

References

Butz, C. J., Hua, S., & Maguire, R. B. (2006). A web-based bayesian intelligent tutoring system for computer programming. Web Intelligence and Agent Systems: An International Journal, 4(1), 77-97. Canali, C., Cardellini, V., Colajanni, M., Lancellotti, R., & Yu, P. (2003). Cooperative Architectures and

Algorithms for Discovery and Transcoding of Multi-version Content. Paper presented at the Eighth

International Workshop on Web Content Caching and Distribution.

(5)

adaptive hypermedia architecture. Paper presented at the fourteenth conference on Hypertext and

Hypermedia, Nottingham, United Kingdom.

De Bra, P., Aroyo, L., & Cristea, A. (2004). Adaptive Web-based educational hypermedia. In M. Levene & A. Poulovassilis (Eds.), Web Dynamics, Adaptive to Change in Content, Size, Topology and Use (pp. 387-410). Heidelberg, Germany Springer.

Felder, R. M. (1993). Reaching the second tier: learning and teaching styles in college science education.

Journal of College Science Teaching, 23(5), 286-290.

Felder, R. M., & Silverman, L. K. (1988). Learning styles and teaching styles in engineering education.

Engineering Education, 78(7), 674-681.

Gertner, A., & VanLehn, K. (2000). Andes: A coached problem solving environment for physics

Intelligent Tutoring Systems (pp. 133).

Henze, N., & Nejdl, W. (2004). A logical characterization of adaptive educational hypermedia. New

Review of Hypermedia and Multimedia, 10(1), 77-113.

Karampiperis, P., & Sampson, D. (2005). Adaptive Learning Resources Sequencing in Educational Hypermedia Systems. Educational Technology & Society, 8(4), 128-147.

Kim, H.M., & Lee, K.H. (2006). Device-independent web browsing based on CC/PP and annotation.

Interacting with Computers, 18(2), 283-303.

Laakko, T., & Hiltunen, T. (2005). Adapting Web content to mobile user agents. IEEE Internet

Computing, 9(2), 46-53.

Noble, B. (2000). System Support for Mobile, Adaptive Applications. IEEE Personal Communications, 44-49.

Papanikolaou, K., Andrew, M., Bull, S., & Grigoriadou, M. (2006). Designing learner-controlled educational interactions based on learning/cognitive style and learner behaviour. Interacting with

Computers, 18(3), 356-384.

Soloman, B. A., & Felder, R. M. (2003). Index of learning styles questionnaire. Retrieved May, 2012, from http://www.engr.ncsu.edu/learningstyles/ilsweb.html.