A Service-Oriented Approach for the Pervasive Learning Grid

JOURNAL OF INFORMATION SCIENCE AND ENGINEERING 21, 959-971 (2005)

A Service-Oriented Approach for the Pervasive Learning Grid CHING-JUNG LIAO, FANG-CHUAN OU YANG* AND KEVIN CHIHCHENG HSU* Department of Management Information Systems Chung Yuan Christian University Chungli, 320 Taiwan E-mail: [email protected] * Department of Information Management National Central University Chungli, 320 Taiwan This work proposes p-Learning grid, a service-oriented approach, based on a pervasive learning grid, for solving difficulties associated with the sharing of learning resources distributed on different e-Learning platforms. The p-Learning grid not only enables collaboration and effective reuse of learning objects but also supports learning anytime, anywhere. Since the WSDL of web services remains poorly defined and has poor dispatch ability in service-level agreements for resource description, distributed resources can not be effectively managed, and service collaboration could not be achieved. Our grid service was generated based on web services and grid technology which support good descriptions of services and management mechanisms. The proposed p-Learning Grid is based on such grid service technologies as Globus Toolkit 3 [28], the Grid Services Flow Language (GSFL) [17], etc., along with mobile devices and relevant technologies for supporting a pervasive and collaborative system in which resources can be effectively managed and shared. This study used three self-developed learning platforms, integrated with GT3, to provide the grid engine used to implement the entire system. The experiment involved the creation of English learning objects accessible via Nokia, Sony Ericsson, and Motorola mobile phones. Keywords: grid services, pervasive learning, p-learning grid, learning objects, GSFL

1. INTRODUCTION Electronic learning (e-Learning) has recently become an important medium of learning, and pervasive learning (p-Learning) refers to learning anytime, anywhere. One of the client devices involves using cellular phones for learning. Several standards for e-Learning currently exist, including IMS [15], SCORM (Sharable Content Object Reference Model) [26], and ULF (Universal Learning Format) [31], which have been combined with XML based technologies to define and describe each e-Learning material as a learning object (LO). The different LOs can be inter-recognized to enable exchanges among different learning systems which support these standards. However, e-Learning systems suffer from several problems. First, e-Learning resources are always distributed around several locations, making it difficult to integrate numerous e-Learning resources. Second, most e-Learning components are system-dependent and cannot be combined Received July 1, 2004; revised February 1, 2005; accepted May 18, 2005. Communicated by Robert Lewis.

959

960

CHING-JUNG LIAO, FANG-CHUAN OU YANG AND KEVIN CHIHCHENG HSU

with other systems. In other words, a component programmed in VB is difficult to migrate to a Unix-like platform and cannot communicate with a component hosted on it. Third, the relationships among LOs cannot be bound; thus, service-level agreements across multiple LO’s are insufficient to control workflow collaboration. Fourth, learners still cannot learn without time and place restrictions. Most e-Learning systems ask learners to use specific client devices to learn. Because of these problems, several researchers have proposed the use of web services to achieve learning object collaboration [8, 9, 16], but WSDL 1.1 [32] remains inadequate due to its inability to integrate status control and workflow at the service level. WSDL 1.1 only defines a service implementation and interface. Consequently, a system for pervasive learning has been devised to solve these problems based on grid service core technologies [29], and it is termed the pervasive learning grid (p-Learning Grid). Grid service core technologies have superior ability to dispatch resources. The purposes of applying grid service core technologies to the p-Learning Grid are as follows: 1. to achieve better collaboration among related LMS’s in order to perform learning tasks that are larger in scope; 2. to provide consistent service-level agreement across various related LMS’s during learning sessions or processes; 3. to provide better adaptive learning across various LMS’s (e.g., a unified scoring/ grading system to allow learners to proceed from LMS 1’s material to the next level of related materials hosted on LMS 2); 4. to provide a learner-centric single portal for learners, which is capable of incorporating all of the learner’s desired learning materials; all of the portal’s capabilities (such as billing, security, etc.) should also be supported in this single portal; 5. to facilitate life long learning, where a learner will need to access many LMS’s during his/her learning career and may wish to have consistent scoring, tracking, accessing, and/or learning methodologies during the life-long learning experience. This study developed a service-oriented solution for the pervasive e-learning system based on a pervasive learning grid using three self-developed learning systems. Ring found that one of the best uses of mobile learning was language learning [25], and since English is the first foreign language in TAIWAN, learning objects for English learning were selected. GT3 was employed as a grid engine for integrating the three proposed learning systems into a pervasive learning grid. Client learners can access the LOs from the pervasive learning grid using a laptop, the iPAQ H3950 PDA, or mobile phones like the Nokia 7210, 6100, and 6610, Sony Ericsson P900, and Motorola 388C. The remainder of this paper is organized as follows: Related works are discussed in section 2. Section 3 then details the proposed system architecture of a pervasive learning grid. Next, section 4 presents and discusses the experimental results. Finally, section 5 gives conclusions and directions for future research.

2. RELATED WORKS The works related to our proposed system are presented here in such a way as to

A SERVICE-ORIENTED APPROACH FOR THE PERVASIVE LEARNING GRID

961

highlight how we leverage existing solutions from Grid technologies and public standards to provide the intended open and interoperable learning systems needed to solve important e-Learning system integration issues. 2.1 Basics of Grid Computing Grid computing focuses on resource sharing and social policies. Resource sharing in grid computing primarily focuses on direct access to computers, software, data, and other resources, as required by various collaborative problem-solving and resource-brokering strategies emerging in industry, science, and engineering [5]. Social policies of grid computing mean that each grid node can have its own hardware and software specifications, storage devices, network topologies, and so on [2]. The word “Grid” often also means “computational grid.” A computational grid is a hardware and software infrastructure that provides dependable, consistent, pervasive, and inexpensive access to high-end computational capabilities [4]. Providing an alternative perspective, Ian Foster [3] considers that grid computing involves “coordinated resource sharing and problem solving in dynamic, multi-institutional virtual organizations.” Virtual Organization (VO) is a set of individuals and/or institutions defined by such sharing rules. These shared resources can include computing resources, data, network bandwidth, and so on. 2.2 Service-Oriented Architecture Currently, grid architectures are shifting toward service-oriented concepts. Grid services with web service technologies are a new development trend and are gradually obtaining enterprise support. A service can be considered a platform-independent software component, which is described using a description language and published as part of a directory or registry by a service provider. A service request can then locate a set of services by querying the registry, a process termed resource discovery. Finally, a suitable service can be selected and invoked, a process called binding. Based on the concept of service-oriented Grid computing, a novel infrastructure was designed by the researchers in GGF [13] and named OGSI (Open Grid Service Infrastructure) [29, 30]. The architecture implemented based on OGSI was named OGSA (Open Grid Service Architecture) [6, 21]. A service that follows the specifications of OGSA can be viewed as a grid service [7]. Sometimes, a grid service can be defined as a service published in the registry in a grid environment. A computational grid comprises a set of resources, such as computers, networks, on-line instruments, data servers, sensors, and so on [12]. From the grid service perspective, the above resources are considered to be services. The grid service that employs the same concept and architecture as the Web service plays the three important roles of service broker, service provider, and service requester. A service broker is a medium between service providers and service requesters. A service broker receives publication requests from service providers and processes service discovery queries from service requesters. Service providers employ WSDL to describe services which have been implemented in the form of a file style and then publish the services to service brokers to provide services to service requesters. Service requesters are individuals who require specific services. Service requesters can ask for services via service brokers and then use URIs (Uniform Resource Identifiers) to bind services from

962

CHING-JUNG LIAO, FANG-CHUAN OU YANG AND KEVIN CHIHCHENG HSU

service providers. The format of messages for communicating among the three roles is SOAP (Simple Object Access Protocol) [27]. Grid services either employ WSDL to describe the service implementation and interface, or provide additional system-level services and mechanisms to manage grid resources. These mechanisms are for dynamically publishing and binding services, managing distributed and heterogeneous grid resources, performing real-time monitoring of the status and performance of each node and service, and establishing service-level agreements related to collaboration among services, security mechanisms, reliability, fault tolerance, and so on. These mechanisms and services can be provided through specific grid containers/engines. Some grid containers, like Globus Toolkit Version 3.2 (GT3) from The Globus Alliance [28], even incorporate WSDL to establish a novel standard- GWSDL (Grid WSDL) [14] in order to improve the descriptive abilities of WSDL1.1. The Grid Services Flow Language (GSFL) was proposed by Krishnan et al. [17] to describe the workflow of different dynamic grid services. GSFL is XML based and solves the problems of peer-to-peer interaction among grid services and lifecycle management of grid services. 2.3 Grid Service Core Technologies for e-Learning Grid services combined with web services technologies are a new trend and are gradually obtaining enterprise support. Based on grid technology and web services, grid services seamlessly gather various heterogeneous, dynamic, and distributed resources from various places, and achieve comprehensive and meaningful sharing of grid resources. Thus, grid services have some advantages over web services in terms of workflow collaboration and resources sharing [17]. Notably, grid services provide a better solution to the problem of learning resource sharing and collaboration for e-Learning [20]. Recently, some researchers have proposed methods for learning resource sharing. For example, Brusilovsky proposed reusable distributed learning activities [1], and Fuji proposed a methodology based on the CORBA technique to make learning objects reusable [10]. Meanwhile, several works have examined how to apply grid service technologies to e-Learning. For example, Reklaitis developed a framework based on Globus and used it to develop a grid environment for e-Learning [22-24]. Gaeta also developed some concepts for employing grid technologies to integrate learning resources [11]. Li et al. designed a continuous pervasive learning system that integrates different types of e-Learning platforms into a pervasive learning environment [18], but they did not use grid service technologies. The main difference between using grid technologies to integrate learning resources and using traditional distributing technologies is that a grid can obtain all the computational information among the grid nodes and, thus, provide multi-dimensional quality of services (QoS) for learning platforms. However, most of the above researchers have only proposed related concepts and ideas, and have not yet presented specific experimental methods and results.

A SERVICE-ORIENTED APPROACH FOR THE PERVASIVE LEARNING GRID

963

3. THE SYSTEM ARCHITECTURE OF THE P-LEARNING GRID This study applied grid service technologies to pervasive learning. GT3 was used to establish an environment for grid services which connected several computers. The LOs distributed among various learning platforms were packed with different types of services, and then these LOs were mapped as standard grid services. The platform was named the pervasive learning grid (p-Learning Grid). Fig. 1 shows the system architecture of the p-Learning Grid. The system can be explained in three parts. The left part of Fig. 1 shows several LO Services supported by different content creators. The LO Services can be either physically located at different positions or hosted in heterogeneous platforms. Each node containing the LO Services represents a virtual organization which can have its own platform architecture, operating system, software, hardware, and organization policy. In the middle of Fig. 1, the service registry enables each LO to register there, thus enabling the service requester to bind services. The difference in the service registry between the p-Learning Grid and web services is such that the service registry of the p-Learning Grid can be any host computer of a node in the grid, but the web services could not reach. Since the p-Learning Grid is constructed based on the grid environment, the host computer of each node is peer-to-peer, and information can be shared and exchanged among all of the nodes. Additionally, the grid core engine monitors the states of each node and registry services to confirm whether or not they are alive. That is, LO services in the p-Learning Grid are dynamically generated, searched, released, and bound. The host of each node in the p-Learning Grid can also be the service registry used for searching services. The right part of Fig. 1 shows the clients of the p-Learning Grid, which can be mobile devices,

Fig. 1. The system architecture of the pervasive learning grid.

964

CHING-JUNG LIAO, FANG-CHUAN OU YANG AND KEVIN CHIHCHENG HSU

for example, tablet PCs, notebooks, PDAs, cellular phones, and so on. The mobile devices can connect to the service registry of the p-Learning Grid to access services via the application interface. The different LOs, the LMS (Learning Management System), and the LCMS (Learning Content Management System) are supported as described in [19]. These LOs focus on English learning, and are packed using the SCORM standard. A Java phone is used for client access in the system implementation. Each VO supports one LO service which can be added or destroyed dynamically. One VO is defined as a service registry and is named the LO service broker. Generally, the LO services broker can be any host in the p-Learning Grid, but a cellular phone must assign a gateway to enter the system by means of GPRS, and then a VO must be identified to perform this task. Meanwhile, grid services are able to tolerate faults, and several hosts can be built to ensure service stability, including VO5 as shown in Fig. 2. VO6 and VO7 are learning platforms designed with PHP based on a Linux system with an Apache server, MySQL. The portal modules are PostNuke and XOOPS, respectively. Moreover, VO8 is a learning platform designed using ASP .NET based on a Windows 2000 Server system with IIS and SQL server, and

Fig. 2. The implementation of the pervasive learning grid.

A SERVICE-ORIENTED APPROACH FOR THE PERVASIVE LEARNING GRID

965

is especially designed to provide content access via mobile phones. In the normal case, the grid container selects the best one from VO5 based on the system performance and network states for the purpose of service binding. When a host failure occurs, the failover mechanism starts to transfer services to the second best host to be executed. In the system operation flow, when clients use cellular phones via the base platform to enter the LO services broker in the p-Learning Grid, the available services can be displayed on the screens of the cellular phones. The LO services broker searches for the requested services in the p-Learning Grid based on the metadata described by the LO services to support the client users. In this study, eight VOs were built, including three different self-developed learning platforms. The number of VOs is scalable in the p-Learning Grid. Generally, the query and registry of learning object services are processed by a broker. Service providers register the services with the broker. The system has three interfaces: a registry, one or more factories, and a mapper to help requesters get the requested services. After the requesters find the desired services using the registry interface, several further processing steps are required for service delivery. First, the grid container executes the authorization service to verify the authority of the user. The resource broker then selects an appropriate host to execute the service. Subsequently, the factory interface instructs the host to generate an application instance. Finally, the binding between the service requestor and the application is finished using the service binding interface. Following these steps, client users obtain the learning content after the completion of the LO service. The messaging technique for publishing and binding services within LO services based on grid service core technologies is similar to the web service method. The most widely-used message technique is SOAP.

4. RESULTS AND DISCUSSION In a pervasive learning grid, GT3 was installed on each node as a grid engine in order to establish grid services. Eight VOs supported different English LO services. In the implementation, VO1, VO2, VO3, and VO4 supported the “One sentence everyday,” “Life English,” “Computer technology English,” and “Business English” services, respectively. The learning records were stored on VO5 to support personalized services. Four hosts established the fault tolerance system at VO5. We also established a gateway (the LO services broker of the p-Learning Grid) at VO5 to allow client users to join the system via laptops, PDA’s, or mobile phones. In the physical implementation, client users could join the system via any node at VO5. The client devices were the Nokia 6100, 6610, and 7210, Sony Ericsson P900, and Motorola 388C mobile phones. The Java J2ME MIDP1.0 specification was used to program the applications, which were tested on emulators supported by Nokia, Sony Ericsson, Motorola, and Sun, respectively. Fig. 3 shows a scenario that we will use to explain the system operations. First, LO services must broadcast to the GT3 server. Client learners then must obtain MIDP mobile phone programs from the proposed platform and install them on the mobile phone. Learners then must input a set of authorized accounts and passwords, so that they can use mobile phones to connect with an LO service broker in the p-Learning Grid (here, the broker is established as a host of VO5 by means of GPRS). LO service brokers list all of the available services. If learners select the service of “One sentence everyday”, the grid

966

CHING-JUNG LIAO, FANG-CHUAN OU YANG AND KEVIN CHIHCHENG HSU

Fig. 3. System operations of the pervasive learning grid.

container (GT3) identifies services by using the supported metadata. Finally, when the learner selects the service from VO1, the GT3 server binds the system services from VO5 automatically to provide the information requested by learners. In this way, VO1 provides proper content to learners based on the learning status and progress. VO1 and VO5 own services themselves, but can also collaboratively support suitable, stable services for learners. The various VOs (from VO1 to VO8) not only are distributed at different places but also are heterogeneous in terms of software and hardware. However, in the p-Learning Grid, the learning resources distributed in different places can integrate and share service types by using the grid core container. At the same time, the services supported by each node can collaborate dynamically; thus, the system can be stabilized with good service quality. To help explain the p-Learning Grid experiment, four learning scenarios were described in the following: Scenario 1 Thomas could interact with his teacher to learn in the multimedia classroom at school. He also could learn by accessing materials from the p-Learning Grid. Scenario 2 When Thomas went home, he could learn by using a PC with access to the p-Learning Grid. He chose several learning materials, e.g., some words, several sentences, and assessments. Finally, the learning record could be customized. Scenario 3 Thomas could log onto the p-Learning Grid via a cellular phone during his vacation. At this time, the p-Learning Grid provided adaptive content by considering his personal learning profile and the device attributes. The adaptive content is shown in Fig. 5.

A SERVICE-ORIENTED APPROACH FOR THE PERVASIVE LEARNING GRID

967

Scenario 4 Thomas used his PDA to learn in the classroom. The p-Learning Grid provided images with better solution and video/audio materials since greater computing power and a superior display where available than could be provided by mobile phones. Figs. 4-6 show the learning contents accessible by a desktop PC, a laptop, a Nokia 7210, and an iPAQ H3950 PDA, respectively. The advantages of the p-Learning grid are the workflow collaboration, content adaptation, and LMS integration.

Fig. 4. One of the learning contents displayed on the laptop or PC.

Fig. 5. A series of learning contents displayed on the Nokia 7210 mobile phone.

968

CHING-JUNG LIAO, FANG-CHUAN OU YANG AND KEVIN CHIHCHENG HSU

Fig. 6. Two of the learning contents displayed on the iPAQ H3950 PDA.

5. CONCLUSIONS This paper has proposed a service-oriented approach to pervasive learning grid to solve difficulties in sharing learning resources distributed on different e-Learning platforms and to help users learn anytime, anywhere. Furthermore, the proposed system produces learning objects that can be used effectively for the purposes of collaboration and reuse. Since web services alone are insufficient to provide good definitions and dispatching capability in the service layer for resource description, the distributed resources are not effectively managed or shared among the services. A grid service framework is generated based on web services and grid technologies. It supports an effective service description and management mechanism. The p-Learning grid based on grid service technologies along with mobile devices and relevant technologies supports pervasive learning. Three self-developed learning systems collaborated by GT3 grid engine and GSFL have been implemented to provide a pervasive learning grid. In our experiment, English learning objects were produced and made accessible to a laptop, an iPAQ H3950 PDA, and to Nokia 7210, 6100, and 6610, Sony Ericsson P900, and Motorola 388C mobile phones. The results of this study demonstrate the effectiveness of the proposed system. Future works will make the p-Learning grid architecture adaptable to 3G (CDMA and WCDMA).

ACKNOWLEDGMENTS The authors would like to thank the NSC of the Republic of China, Taiwan, for financially supporting this research under Contract No. NSC 92-2213-E-033-026.

A SERVICE-ORIENTED APPROACH FOR THE PERVASIVE LEARNING GRID

969

REFERENCES 1. P. Brusilovsky and H. Nijhavan, “A framework for adaptive e-learning based on distributed reusable learning activities,” in Proceedings of the World Conference on E-Learning in Corporate, Government, Healthcare, and Higher Education, 2002, pp. 154-161. 2. I. Foster, “The grid: a new infrastructure for 21st century science,” Physics Today, Vol. 55, 2002, pp. 42-47. 3. I. Foster, “What is the grid? A three point checklist,” 2002, http://www-fp.mcs.anl. gov/~foster/Articles/WhatIsTheGrid.pdf. 4. I. Foster and C. Kesselman, “Computational grids,” Chapter 2, in The Grid: Blueprint for a New Computing Infrastructure, Morgan- Kaufman, 1999. 5. I. Foster, C. Kesselman, and S. Tuecke, “The anatomy of the grid enabling scalable virtual organizations,” International Journal of Supercomputer Applications, Vol. 15, 2001, pp. 200-222. 6. I. Foster, C. Kesselman, J. Nick, and S. Tuecke, “The physiology of the grid: an open grid services architecture for distributed systems integration,” Open Grid Service Infrastructure WG, Global Grid Forum, 2002, http://www.globus.org/research/ papers/ogsa.pdf. 7. I. Foster, C. Kesselman, J. Nick, and S. Tuecke, “Grid services for distributed system integration,” Computer, Vol. 35, 2002, pp. 37-46. 8. G. Fox, “Education and the enterprise with the grid,” Chapter 43, in Grid Computing: Making the Global Infrastructure a Reality, F. Berman, G. Fox, and T. Hey (eds.), John Wiley and Sons Ltd., Chichester, 2003. 9. G. Fox, S. H. Ko, K. Kim, S. Oh, and S. Lee, “Integration of hand-held devices into collaborative environments,” in Proceedings of 2002 International Conference on Internet Computing (IC-02), 2002, pp. 231-250. 10. T. Fuji and T. Tanigawa, “The methodology for reuse of e-learning resources,” in Proceedings of the World Conference on E-Learning in Corporate, Government, Healthcare, and Higher Education, 2002, pp. 305-310. 11. M. Gaeta, P. Ritrovato, and S. Salerno, “Implementing new advanced learning scenarios through grid technologies,” in Proceedings of 1st LeGE-WG International Workshop on Educational Models for GRID Based Services, 2002, http://ewic.bcs. org/conferences/2002/1stlege/session1/paper3.htm. 12. D. Gannon and R. Bramley, et al., “Programming the grid: distributed software components, p2p and grid web services for scientific applications,” Cluster Computing, Vol. 5, 2002, pp. 325-336. 13. Global Grid Forum, http://www.gridforum.org/. 14. GWSDL, http://www-unix.globus.org/toolkit/3.0/ogsa/impl/java/build/javadocs/org/ globus/ogsa/ samples/gwsdl/top/bindings/package-frame.html. 15. IMS, http://www.imsproject.org. 16. K. Krauter, R. Buyya, and M. Maheswaran, “A taxonomy and survey of grid resource management systems for distributed computing,” Software: Practice and Experience, Vol. 32, 2002, pp. 135-164. 17. S. Krishnan, P. Wagstrom, and G. Laszewski, “GSFL: a workflow framework for grid services,” 2002, http://www-unix.globus.org/cog/papers/gsfl-paper.pdf.

970

CHING-JUNG LIAO, FANG-CHUAN OU YANG AND KEVIN CHIHCHENG HSU

18. L. Li, Y. Zheng, H. Ogata, and Y. Yano, “Research on pervasive e-learning system development,” in Proceedings of World Conference on E-Learning in Corporate, Government, Healthcare, and Higher Education, 2003, pp. 595-598. 19. C. J. Liao, “A WWW-based e-learning problem-solving environment,” in Proceedings of Conference on Information to Apply and Develop, 2002, pp. 90-95. 20. C. J. Liao, “The collaboration of learning objects in a grid environment,” IEEE Learning Technology, Vol. 6, 2004, pp. 19-22. 21. OGSA Working Group, https://forge.gridforum.org/projects/ogsa-wg. 22. V. Reklaitis, K. Baniulis, and A. Masevicius, “Towards e-learning application architecture based on globus framework,” in Proceedings of Euroweb 2002 Conference, 2002, http://www.w3c.rl.ac.uk/Euroweb/poster/116/Kaunas.html. 23. V. Reklaitis, K. Baniulis, and T. Okamoto, “Shaping e-learning applications for a service-oriented grid,” in Proceedings of 2nd International LeGE-WG Workshop on e-Learning and Grid Technologies: a Fundamental Challenge for Europe, 2003, http://ewic.bcs.org/conferences/2003/2ndlege/session1/paper1.htm. 24. V. Reklaitis, “Towards e-learning grid services,” in Proceedings of World Conference on E-Learning in Corporate, Government, Healthcare, and Higher Education, 2003, pp. 692-695. 25. G. Ring, “Case study: combining web and wap to deliver e-learning,” http://www. learningcircuits.org/2001/jun2001/ring.html, 2001. 26. SCORM, http://www.adlnet.org. 27. Simple Object Access Protocol (SOAP) version 1.1, http://www.w3.org/TR/SOAP/. 28. The Globus Alliance, http://www.globus.org/. 29. S. Tuecke, et al., Grid Service Specification, http://www.gridforum.org/ogsi-wg/ drafts/draft-ggf-ogsi-Gridservice-05_2002-11-04.pdf, 2002. 30. S. Tuecke, K. Czajkowski, I. Foster, J. Frey, S. Graham, C. Kesselman, T. Maguire, T. Sandholm, P. Vanderbilt, and D. Snelling, Open Grid Services Infrastructure (OGSI) Version 1.0. Global Grid Forum Draft Recommendation, 2003. 31. ULF, http://www.saba.com. 32. Web Services Description Language (WSDL) version 1.1, http://www.w3.org/TR/ wsdl.

Ching-Jung Liao (廖慶榮) is an Assistant Professor in Department of Management Information Systems at Chung Yuan Christian University, Taiwan. He received his B.S. and M.S. degrees in Computer Science and Biomedical Engineering from Chung Yuan Christian University, respectively, and a Ph.D. of Information Engineering and Computer Science from Feng Chia University. He has been a guest scientist in the Institute of Informatics of Technical University of Munich, Germany. His current research interests include grid computing, parallel and distributed computing, e-learning, pervasive learning, and ubiquitous computing.

A SERVICE-ORIENTED APPROACH FOR THE PERVASIVE LEARNING GRID

971

Fang-Chuan Ou Yang (歐陽芳泉) received his B.S. and M.S. degrees in Department of Management Information Systems at Chung Yuan Christian University, Taiwan. He is currently a Ph.D. student in Department of Information Management at National Central University, Taiwan. His research interests include e-learning system integration, grid computing, distributed systems, Web services, and web portal technologies.

Kevin Chihcheng Hsu (許智誠) is an Assistant Professor in Department of Information Management at National Central University, Taiwan. He received his B.S. of Computer Science from National Taiwan University, a M.S. of Computer Science from State University of New York at Stony Brook, and a Ph.D. of Computer Science from University of California, Los Angeles. He has extensive industry experience in distributed systems, workflow, B2B, multi-tier web application, and system integration. His current research interests include e-learning system integration, workflow learning in enterprise, business process management, and multi-tier web applications within supply chain.