Personalized Service-Oriented E-Learning Environments

62 62 Madrid of University III Carlos Kloos Delgado Carlos and Muñoz-Merino, J. Pedro Muñoz-Organero, Mario ublished by the IEEE Computer Society Computer IEEE the by ublished E-Learning P E-Learning Environments Personalized Service-Oriented

T environments. learning (student-led) informal e-learning in and (instructor-led) formal both of needs trends the captures that architecture new current analyze a describe They authors also limitations. and merits identify and evolution architecture The systems. e-learning Consequently, next-generation of wave latest environment. the define to changing starting already are rapidly researchers a in to needs adapt learning must and life-long offer customization require many they environments because users study Learning case ideal offers an personalization. services interface 2.0-related for Web opportunities of component social The ligent tutoring systems. tutoring ligent ferent users’ heterogeneous needs by by hypermedia adaptive needs using heterogeneous users’ ferent dif account In into took they (http://black pattern. particular, this Blackboard followed board.com) (http://dotlrn. and org), dotLRN moodle.org), (http:// Moodle as such LMSs implementa of tions former — environments learning centralized stan dard-supporting, component-based, LMSs into evolved platforms, monolithic prietary pro From environment. learning tual vir a called learning-management sometimes (LMS), system central a knowledge via instructor-provided uting early 1990s were based on distrib based were 1990s early the in platforms e-learning viding pro and defining in steps first he 1 089-7801/10/$26.00 ©2010 IEEE 089-7801/10/$26.00 2

1 and intel and

------

ltom wl apy evc frame service support design, modular to works their apply will e-learning platforms next-generation to example, for colleagues, his and According Dagger Declan system. management central a use they whether but design modular their isn’t them between ence differ main apply the to PLEs; also can but LMSs to restricted aren’t concepts service-oriented are that in nonorthogonal families Both (PLE). environment learning personal a of propose architecture that the those and service- a LMS oriented define that two in those this trends families: these In place we architectures. article, e-learning approaches to new disruptive some sometimes and defined have researchers However, in the past seven years, years, seven past the in However, I EEE INTERNET COMPUTING INTERNET EEE - - - Personalized Service-Oriented E-Learning Environments

ing the idea of an LMS composed by integrating okiproject.org) defines service layers for devel- interoperable services.3 oping e-learning platforms. All these efforts However, Scott Wilson and his colleagues decompose the different functionalities in a focus on PLEs, letting users apply Web 2.0-style traditional LMS into a set of services and iden- services to create their own learning manage- tify the methods required for a distributed ment tools.4 Charles Severance and his col- interface to access such functionality. They leagues combine the concepts of both families, also subdivide services into several catego- applying PLE-style approaches to a mashup- ries, including basic services provided by the based LMS — here, the LMS shrinks to become e-learning infrastructure (such as HTTP), com- a much simpler container that joins and orga- mon services required in all e-learning envi- nizes the capabilities of a wide range of tools ronments (such as authentication, file sharing, around a particular learning context.5 This or logging), and specific application services combination of concepts in both families relies (such as quizzes or simulations). on learner-generated contexts (http://learner- The service-oriented LMS offers many impor- generatedcontexts.pbwiki.com) as well as peer- tant benefits compared with the conventional to-peer learning management and groupware monolithic LMS. For example, it provides a more systems such as Colloquia (www.colloquia.net). flexible architecture in which instructors can Here, we present an additional way to add and use new services dynamically. Instruc- combine the benefits from centralized service-oriented LMSs and PLEs by defining a service-oriented personalized e-learning envi- Students will no longer passively ronment. We also identify and define some open issues, along with the results of our proposed consume learning materials but actively architecture’s implementation. create and disseminate knowledge. Current E-Learning Trends As we mentioned earlier, one of the first approaches to a next-generation e-learning tors can also integrate outside expert knowledge architecture decomposed a central LMS into a into courses offered on that platform via distrib- service-oriented e-learning platform.3 Dagger uted interfaces; alternatively, system developers and his colleagues believe that separating LMS can reuse code across every learning platform and learning-content-management-system because of the code’s distributed nature. (LCMS) functionality provides support for Another approach we mentioned ear- greater interoperability, in which systems not lier defines a PLE in which every user builds only share content and learning scenarios but his or her own learning path by using avail- also exchange tools, functionalities, seman- able services on the Internet. Downes expects tics, and control seamlessly and dynamically. to see more research examining the concept of This definition disconnects new, innovative “e-learning 2.0,” in which users apply popular services and their applications to e-learning social networking tools to e-learning processes.7 from a particular central LMS technology. In The traditional asymmetric learning environ- fact, in this service-oriented architecture, a ment, with its clear distinction between the system developer can implement a new LMS roles of instructors and students, will become platform simply by using a new set of differ- more symmetric and based on communities of ent, open e-learning services.6 practice. Students will no longer passively con- The standards community has made some sume learning materials but actively create and strides in defining a service-oriented e-learn- disseminate knowledge. PLEs emphasize sym- ing platform. The IMS Abstract Framework metric connections with a range of services (www.imsglobal.org/specifications.html) iden- in both formal (instructor-led) and informal tifies and represents the core components and (student-led) learning, work, and leisure. Rather interfaces. The E-Learning Framework (ELF; than integrating tools within a single context, www.elframework.org) illustrates e-learning PLEs coordinate connections between users and systems’ common functionalities; similarly, a wide range of services offered by organiza- the Open Knowledge Initiative (OKI; www. tions and other individuals.

MARCH/APRIL 2010 63 E-Learning

External service Service components collaborative interactions.9 Our proposed architecture allows both user- and instructor-controlled learning processes. Using com- PLE munity-defined units of learning isn’t a new concept,9 but our architecture enhances an IMS-LD (www.imsglobal.org/learningdesign/) IMS-LD player Personal service controllable learning process with the repro- User of instructor-dened UoLs Personal service execution environment duction of instructor-, community-, or user- defined units of learning that might contain pluggable external services. Figure 1. A service-oriented personalized e-learning environment. The user in the center consumes external e-learning services Defining a orchestrated by an IMS-LD. Service-Oriented Environment Ultimately, next-generation educational systems should emphasize the use of external Personal e-learning environments are well services and be adaptable for both formal and adapted to the life-long learning needs of our informal learning. They should also capture current IT-based society. They provide the mechanisms to define competencies and pro- required flexibility each user wants, especially vide community-agreed orchestrated paths for as users adapt tools to a particular context. acquiring them. Finally, they should separate Life-long learning requires users to develop the central orchestration point or LMS from the their own competences, so the TENCompetence external e-learning services available on the consortium is developing a domain model to Internet.7 In this section, we describe how our address this need.8 For specific courses, adaptive service-oriented personalized e-learning envi- hypermedia systems1 and intelligent tutoring ronment tries to meet all these goals. Figure 1 systems2 offer alternative approaches by devel- shows our proposed architecture. oping systems that can adapt to individual user The central element is the user, who runs a goals, tasks, and interests. Integrating intelli- PLE to access external e-learning services. This gent tutoring systems with personal e-learning PLE is actually an aggregator of different ser- environments might combine short- and long- vices available from different communities.5 term user adaptation (the EU’s Adaptive Learn- These services can be independently provided ing Spaces project [ALS; www.als-project.org] either by the user (personal e-learning services) has done some preliminary research). or third parties. They can also be orchestrated Although related to the PLE concept, a by a central coordinator tasked with providing third approach incorporates learner-generated user-, community-, or instructor-defined learn- contexts and peer-to-peer systems to share ing designs. Our architecture uses IMS-LD to resources. Hubert Vogten and his colleagues define these learning designs. Using IMS-LD, developed an implementation that follows our architecture can orchestrate any service as these principles.9 long as it uses a Web service-based interface. Although learning institutions involved in To provide a modular architecture, our proposal formal learning courses might prefer the mer- decomposes e-learning services into service- its of a centrally controlled service-oriented reusable components.3 LMS, time-constrained users might find PLEs Each service in Figure 1 follows a modu- and learner-generated contexts more flex- lar architecture that defines two independent ible and adaptable to their needs. Dagger and interfaces for the service.10 The user-oriented his colleagues explain that service-oriented interface provides direct service access and LMSs can still benefit from the availability user-to-user interaction, which simplifies cre- of new open e-learning services, but Wil- ation of ad hoc learning communities. The pro- son and his colleagues leave LMS needs out grammatic interface creates a path for service of the e-learning picture. Although PLEs can integration in a common player tool that the plug e-learning services into a learner-con- user can control or that an instructor-defined trolled environment,5 it’s important to provide learning path can centrally manage. The archi- flexible mechanisms for learner-to-learner tecture described in an earlier paper is modular

64 www.computer.org/internet/ IEEE INTERNET COMPUTING Personalized Service-Oriented E-Learning Environments

IMS-LD_ServiceProperties

Email_ServiceProperties Conference_ServiceProperties . . .

(a) Async-conf_ServiceProperties Sync-conf_ServiceProperties

tns:Loc-properties +

tns:Locals – – tns:Locrole-properties +

tns:Locpers-properties +

tns:Glob-properties + tns:Globals – – IMS-LD_ServiceProperties – – tns:Globpers-properties +

tns:nished

tns:visible

tns:Roles – – tns:Roles + 0 .. ∞ (b) any Figure 2. E-learning service hierarchy and resource properties document. (a) E-learning services are organized in a hierarchical tree of services. The figure shows some of the basic e-learning services in that tree. (b) The state of each e-learning service is captured in a resource properties document. The figure captures the state-dependent data of the generic e-learning service on top of the hierarchy. in that it separates the common functionalities creation of pedagogically enhanced platforms required by different services into distributed via IMS-LD’s Web service-based interface, service components.10 Defining service com- which makes e-learning services pluggable to ponents in service-oriented architectures is a external LMS-based or user-controlled players. common practice on the Internet (OpenID [www. Implementing an IMS-LD-compatible interface openid.org] and the del.icio.us API [http://del. also makes it possible to integrate learning ser- icio.us/help/api/] offer two examples) and in vices into formal learning processes. e-learning in particular. Defining this programmatic interface requires Figure 1 also shows our architecture’s con- solving how to synchronize the execution state cept of a personal service execution environ- of each e-learning service and its representa- ment (PSEE), which is associated with the tion inside the IMS-LD player and how to define execution of e-learning services on personal and implement the methods the interface sup- mobile devices. Through their mobile devices, ports. We can categorize the states defined by learners become pervasive consumers of dis- an IMS-LD learning unit instance as tributed learning resources and play an active part in the e-learning scenario, offering addi- • level-B properties, tional resources to other learners. Combining • implicit level-A properties, these devices either in a predefined or ad hoc • service initialization parameters, and manner, we can create personalized distributed • execution environment-dependent states learning architectures. (such as the roles a particular user plays). We designed the architecture in Figure 1 for life-long learning needs. Different users can join Among the different Web-related technolo- available Web 2.0 communities yet still control gies for capturing state-dependent data,11 we how they access and even execute e-learning selected resource properties as defined in the services. Moreover, the architecture allows the Web Service Resource Framework (WSRF).12 To

MARCH/APRIL 2010 65 E-Learning

WSProcessorServlet UIServlets vice; the second captures callback methods for the reverse path. Our service-oriented, pluggable architecture is designed to fulfill the Servlet engine: HTTP request HttpServletRequest, requirements of both formal learning institu- HttpServletResponse, tions and learners. HTTP response Cookies Implementing a Figure 3. Simplified architecture of our implementation of an Personalized E-Learning Environment e-learning service. The user consumes the e-learning services using Pervasive personal e-learning environments a servlet-based HTTP interface controlled by a servlet engine. The should provide an anytime–anywhere learning programmatic access to the services implements a Web services scenario in which mobile personal devices play engine on top of the servlet engine. an important role. Our architecture encourages e-learning users to contribute to already estab- lished e-learning communities by sharing services such as blogs, personal files, and personal forums or syndicated channels running on their mobile phones and PDAs. To integrate mobile services into externally defined units of learning, we need to implement the interfaces described in the previous section, which requires mobile devices to have Web service server capabilities. David C. Chu and M. Humphrey propose an implementation called OSGI.NET, a middleware layer for stateful Web services for Windows mobile Figure 4. User interface for a forum service. The user interface is devices.13 However, this implementation based on HTTP. The figure shows the visualization of a service on isn’t pervasive enough because it’s limited several browsers running on different devices. to devices supporting a particular operating system. Similarly, another implementation described elsewhere is limited to Symbian provide a generic yet scalable architecture, we devices.14 To provide a non-operating, system- defined a hierarchical structure of e-learning dependent implementation, we defined and services that our implementation translates into implemented Web service development mid- a hierarchical structure of resource properties. dleware in the J2ME MIDP profile (www.jcp. Figure 2 captures part of the service hier- org/en/jsr/detail?id-271), which is based on a archy. Every new service can benefit from simplified servlet API implementation. existing services by its inclusion in the hierar- To validate our architecture’s feasibility chy tree. Figure 2 also shows the root service’s in general and deployment on limited mobile generic resource properties document, which devices in particular, we implemented several captures the state that all IMS-LD synchroniz- personal e-learning services, including a per- able e-learning services share. Essentially, it sonal forum. We divided the forum’s architec- contains both explicit and implicit properties as ture into three parts: the first maintains the defined in a generic learning design. dialogue with the user over a servlet-based Another issue involves how to provide graphical interface, the second is the implemen- e-learning services with the required capa- tation of the IMS-LD Web service, and the third bilities for their inclusion in formal learning is a service component for authentication based processes, which requires defining and imple- on the Open ID specification (http://openid. menting an IMS-LD programmatic interface. net/). Figure 3 shows a simplified architecture The WSRF defines Web service-based distrib- of the application. uted middleware technology; accordingly, our Mobile devices often have HTML-based architecture defines two WSRF-based inter- browsers, some of which are commercial and faces. The first captures communication needs some of which are open source. Examples from the IMS-LD player to the distributed ser- include Bitstream’s Thunderhawk client (www.

66 www.computer.org/internet/ IEEE INTERNET COMPUTING Personalized Service-Oriented E-Learning Environments

bitstream.com/wireless/), Mozilla’s Minimo Web 9. H. Vogten et al., “Using the Personal Competence Man- browser (www.mozilla.org/projects/minimo/), ager as a Complementary Approach to IMS Learning and the Opera Mini browser (www.operamini. Design Authoring,” Interactive Learning Environments, com). Figure 4 shows our forum service visual- vol. 16, no. 1, 2008, pp. 83–100. ized in an Opera Mini browser on a simulated 10. M. Muñoz-Organero and C. Delgado Kloos, “Perva- Nokia 6131 Near-Field Communication mobile sive LMS Components and Services for a Pervasive device, in a Dell Axim X50v PDA running a M-Learning Architecture,” Int’l J. Mobile Learning and Pocket Internet Explorer, and in a Nokia E61i Organisation, Oct. 2007, pp. 275–287. running an Opera Mini browser. The forum ser- 11. M.L. Bote Lorenzo et al., “Tailorable Collaborative vice is also executed in a mobile device run- Learning System that Combines OGSA Grid Services ning J2ME. and IMS-LD Scripting,” Proc. 10th Int’l Workshop Groupware, LNCS 3198, Springer, 2004, pp. 305–321. 12. T. Banks, Web-Services Resource Framework (WSRF): he architecture we described here is open and Primer, v. 1.2, Oasis, May 2006; http://docs.oasis-open. T service-oriented, enabling the integration of org/wsrf/wsrf-primer-1.2-primer-cd-02.pdf. existing learning services, especially those built 13. D. Chu and M. Humphrey, “Mobile OGSI.NET: Grid with Web 2.0 features and functionality. Here, Computing on Mobile Devices,” Proc. 5th IEEE/ACM we described its implementation with a sample Int’l Workshop Grid Computing (GRID 04), ACM Press, pervasive e-learning service; we plan to run a 2004, pp. 182–191. trial with real users in the next year. 14. D. Rocha, “Enterprise and Web-Service Programming for the Nokia Smartphone Platforms,” Forum Nokia Acknowledgments Americas, 2006. Spain’s Programa Nacional de Tecnologías de la Sociedad de la Información supported this research through projects Mario Muñoz-Organero is an associate professor of tele- TSI2005-08225-C07-01 and -02. matics engineering at the Carlos III University of Madrid. His research interests include open architec- References tures for e-learning systems, open service creation 1. P. Brusilovsky and M.T. Maybury, “From Adaptive environments for next-generation networks, advanced Hypermedia to Adaptive Web,” Comm. ACM, vol. 45, mobile communication systems, and pervasive com- no. 5, 2002, pp. 31–33 puting. Muñoz-Organero has a PhD in telecommuni- 2. T. Murray, “Authoring Intelligent Tutoring Systems: An cations engineering from the Carlos III University of Analysis of the State of the Art,” Int’l J. Artificial Intel- Madrid. Contact him at [email protected]. ligence in Education, vol. 10, 1999, pp. 98–129. 3. D. Dagger et al., “Service-Oriented E-Learning Plat- Pedro J. Muñoz-Merino is a visiting professor of telematics forms: From Monolithic Systems to Flexible Ser- engineering at the Carlos III University of Madrid. His vices,” IEEE Internet Computing, vol. 11, no. 3, 2007, main research areas are e-learning and the Semantic pp. 28–35. Web. Muñoz-Merino has a PhD in telematics engineer- 4. S. Wilson et al., “Personal Learning Environments: ing from the Carlos III University of Madrid. Contact Challenging the Dominant Design of Educational him at [email protected]. Systems,” Proc. European Conf. Technology Enhanced Learning (EC-TEL 06) Workshops, 2006, pp. 173–182. Carlos Delgado Kloos is a full professor of telematics engi- 5. C. Severance, J. Hardin, and A. Whyte, “The Coming neering at the Carlos III University of Madrid, where Functionality Mash-Up in Personal Learning Environ- he’s also associate vice chancellor, director of two ments,” Interactive Learning Environments, vol. 16, no. master’s programs (one on e-learning), and director of 1, 2008, pp. 47–62. the Nokia Chair. His main research interest is educa- 6. P. Westerkamp, Flexible E-Learning Platforms: A Ser- tional technologies. Delgado Kloos has a PhD in com- vice-Oriented Approach, Logos Verlag, 2006. puter science from the Technical University of Munich 7. S. Downes, “E-Learning 2.0,” e-Learn Magazine, 16 Oct. and a PhD in telecommunication engineering from the 2005, p. 1. Technical University of Madrid. Contact him at cdk@ 8. R. Koper and M. Specht, “Tencompetence: Lifelong it.uc3m.es. Competence Development and Learning,” Competencies in Organizational E-Learning: Concepts and Tools, M. Selected CS articles and columns are also available Sicilia, ed., Idea Group, 2006, pp. 230–247. for free at http://ComputingNow.computer.org.

MARCH/APRIL 2010 67