The Four Elements of a viable PLE Sandy El Helou Denis Gillet Swiss Federal Institute of Technology in Lausanne Swiss Federal Institute of Technology in Lausanne (EPFL) (EPFL) 1015 Lausanne 1015 Lausanne Switzerland Switzerland sandy.elhelou@epfl.ch denis.gillet@epfl.ch ABSTRACT To better address the requirements of lifelong learning, In this paper, we propose and discuss four fitness features educational systems need to become part of an external system considered as essential for developing personal learning accounting for learning inside and outside formal academic environments (PLE) that are viable and ready for appropriation. environments [8]. There is a need to shift from traditional LMS applications particularly focused on formal interactions and Categories and Subject Descriptors learning, to online personal learning environments (PLE) K. [Computing Milieux]: K.3. Computer and Education, K.4 supporting both institutional and self-directed, intended and Computers and Society- miscenallenous. accidental learning. Successfully sustaining lifelong learning with online PLE requires developing and adopting new design patterns, models, and prototypes that can substitute for prevalent General Terms LMS design patterns [9]. In this paper, we discuss four elements Design, Human Factors deemed important for ensuring an online PLE’s fitness for adoption and lifelong survival. Keywords Personal Learning Environments, lifelong learning, knowledge management, social media 2. THE FOUR PLE ELEMENTS This paper is based on the following definition of online PLEs: 1. INTRODUCTION online PLEs are environments that are built from the perspective Rather than being confined to earlier life stages and strictly of the individual rather than the institution [10] and give learners acquired within standard educational systems, learning should the opportunity to decide their own learning goals, control their be actively pursued during the lifetime period. “Lifelong, learning spaces [11] and interact with each other during the lifewide, voluntary, and self-motivated” learning [1] refers to the learning process [12]. The four identified features for building activities that people conduct during their lifetime, to develop successful PLE are described below. knowledge and competences, motivated by personal, social as well as employment reasons [2,3]. Lifelong learning is about 2.1 Encouraging active participation by learning anything, anywhere, anytime and anyway. It adopting social media paradigms encompasses formal, non-formal and informal learning. Formal The problem of low participation and lack of personal incentives learning refers to intentional learning that occurs in structured was a major issue in early collaborative applications [13]. By contexts, and often leads to a formal recognition (e.g. diploma, adopting a user-centered bottom-up philosophy and relying on certificate). Non-formal and informal learning, on the other Web 2.0 technologies, social media applications have hand, take place in environments that are neither essentially successfully overcome several problems identified by earlier learning-oriented, nor structured in terms of learning objectives, CSCW studies, achieving by that a higher acceptability and a material, time, or support [4]. Different from non-formal better user experience than traditional groupware. Online PLE learning, informal learning is accidental or spontaneous, and should embrace the social media practices of knowledge occurs over the lifetime period [5,6]. “democratization” encouraging active participation and Traditional LMS (Learning Management Systems) are not facilitating information dissemination as well as social suitable for lifelong learning. LMS systems are usually interactions. characterized by a hardcoded asymmetry in user rights [7]. First, having low learning curves and offering interactive Students usually have single predetermined roles, share the same user-friendly interfaces is crucial for achieving fitness. With homogenous learning context, and are expected to achieve the respect to developing interactive interfaces and improving the same learning goals within the same period. Moreover, learning content is pre-packaged in learning units, has a restricted visibility scope (usually limited to the course duration), and is isolated from the outside world. Sometimes, courses cannot even be shared within the same LMS. 30 user experience, Web 2.0 technologies such as AJAX1 play a people in different contexts. At the same time, it should not be particularly important role if applied properly [14]. imposed on learners to explicitly specify their interaction and Second, PLE should encourage learner-generated content by learning contexts. PLE should allow different ways of context providing easy individual and collaborative authoring features identification, ranging from those explicitly delimited by such as blogs and wikis. Learner-generated metadata can be learners to those implied from their personal and collaborative achieved by offering social tagging. The term folksonomy actions. On the one hand, a community space constitutes an denotes the Web 2.0 way of organizing content using tags explicit context for potential interactions and learning revolving created and shared by people [15]. around the community’s practices and involving its members, its shared artefacts, as well as its eventual sub-activity spaces. On Third, PLE should combine content management facilities with the other hand, two or more actors commenting the same asset social networking features allowing people to explicitly build could also form an implicit interaction context involving them, and publish their own network of connections. People achieve the asset in question, its owner, and other contributors. lifelong learning by creating, maintaining, extending and Identifying interaction and learning contexts is crucial in PLE strengthening their personal network composed of people with and is indeed more challenging than in traditional LMS. This is similar interest, groups, systems and specialized information sets mainly because PLE are not confined to preplanned [16]. collaborative scenarios occurring within rigid and closed Fourth, PLE should incorporate SALT features. SALT (Share, collaboration spaces. Instead, it also accounts for smoother Assess, Link, Tag) is an acronym introduced in [17] to account forms of interactions that can evolve over time and induce both for social media features that facilitate information intended and unintended learning situations. dissemination and trigger interactions and reflection on knowledge artefacts. Assessment includes liking/disliking, 2.3 Offering elastic community and content commenting, and rating. Giving users the opportunity to easily management services contribute and express their views leads to a better appropriation Communities of practice (CoPs) are defined as a group of of the online platform and increases their motivation to individuals who choose to collaborate on a regular basis in order collaborate with others. Creating links (or bookmarks) to people to learn and improve their practices related to a shared passion and content and sharing them allows discovering the or topic of interest [22]. CoPs are considered to play a key role connections between different items, and discovering new items in fostering knowledge sharing and learning [23]. This triggers through their connections with known ones. Tagging can be the motivation to sustain the initiation and evolution of CoPs in used for describing an item or categorizing it using a professional and educational environments [24]. When it comes user-defined label. Additionally, using tag-based search and tag to groupware systems, flexibility is a critical usability factor and clouds, learners can discover communities, activities, and their design should take into account the possibility for groups to artefacts that are relevant to specific topics of interest. Tagging evolve over time in terms of behavior, nature, and composition people have also proven to be useful in formal contexts [18]. [25]. The same should apply for the support of community Influenced by users’ tagging practices in collaborative tagging building and evolution in a PLE. Users enter their PLE as systems, tag semantics can emerge and evolve [19]. This helps individual actors and not as pre-labeled members of a rigid communities to incrementally build a common vocabulary and organizational or institutional structure. Then, they can create externalize their shared memory. A direct advantage of their self-organized communities [26] or deliberately join incorporating these social media features is generating existing ones, some of which may correspond to institutions and unobtrusive relation-based recommendations whereby metadata organizations. With respect to rights managements, there ought resulting from SALT actions are exploited in order to bring to the to be no pre-assumed hierarchy or default distribution of rights; surface relevant people, activities, and knowledge artefacts a person can be a learner in one community and a moderator in based on how and by whom they have been “salted”. another. 2.2 Representing interaction and learning With respect to content management, learners should be able to create, share, modify, annotate, review and most importantly contexts in a flexible way repurpose learning artefacts ranging from books to Weblogs, Ackerman identifies the necessity of providing flexible, nuanced videos, podcasts and discussion archives [27]. Bringing together and contextualized CSCW (Computer-Supported Collaborative heterogeneous information sources requires adopting Work) apparatus just as human behavior is “flexible, nuanced lightweight specifications such as RSS (Real Simple and contextualized” [20]. This statement perfectly applies to Syndication or Rich Site Summary) [28] and creative commons2 PLE that should be designed in a flexible and bottom-up way licenses rather than strictly adhering to educational standards and account for heterogeneous interaction and learning (i.e. IMS3, SCORM4). Unlike traditional LMS where knowledge contexts, including work, formal learning, and even play [21]. objects are organized within learning units and their usage Learners should be given the opportunity to design and manage anticipated, in a PLE, artefacts can exist outside the scope of their own learning “contexts” by mashing up application activity spaces; they can be shared directly among actors widgets and useful artefacts, then sharing them with different without having to belong to an activity space or fall under the 2 1 AJAX (Asynchronous JavaScript and XML) combined http://creativecommons.org 3 technologies exchange data asynchronously with the server to http://www.imsglobal.org respond to a user’s request. This avoids freezing the current 4 http://en.wikipedia.org/wiki/Sharable_Content_Object_Referen ce_Model 31 umbrella of reaching an explicitly stated objective. Indeed, they [6] Faure, E., Herrera, F., Kaddoura, A. R., Lopes, H., can at any time be posted in one or more activity spaces, Petrovski, A. V., Rahnema, M., et al. (1972). Learning to grouped together in a bottom-up way using tags, or explicitly Be. Paris: Unesco. related to other artefacts. This approach increases the learning [7] Wilson, S., Liber, P. O., Johnson, M., Beauvoir, P., & flexibility and encourages the spontaneous appropriation of Sharples, P. (2007). Personal Learning Environments: knowledge artefacts. Challenging the dominant design of educational systems. Journal of e-Learning and Knowledge Society , 3 (2), 27- 2.4 Providing personalized and contextual 28. Retrieved 2010 20-January from recommendation services http://je-lks.maieutiche.economia.unitn.it/index.php/Je-LK PLE can be classified as “open corpus” environments [29]. In a S_EN/article/viewFile/247/229 PLE, relationships between knowledge artefacts are not [8] Taylor, F. P. (2004). Education technology helps unite necessarily known beforehand, as it is the case in traditional school communities, improve academic achievements. hypermedia systems; instead, they can emerge, evolve, and T.H.E. Journal , 31 (10), 46–48. expand during run time. In addition, in online platforms where everyone is a “consumer” and a “producer”, contributions differ [9] Downes, S. (2010). New Technology Supporting Informal in quality, style, subject matter, target audience, composition, Learning. Journal of Emerging Technologies in Web and reliability. In such open environments, personalized and Intelligence , 2 (1), 7. contextualized recommendations can drive learners’ attention to [10] Atwell, G. (2006 1-June). Wales Wide Web. Retrieved 2010 potentially interesting resources depending on their implicit or 1-May from Personal Learning Environments: explicit interests, therefore avoiding information overload, and http://www.knownet.com/writing/weblogs/Graham_Attwell triggering formal and informal learning opportunities [30,31]. /weblog.archives/2006/6 As mentioned earlier, PLE-embedded recommender systems can exploit SALT actions performed by users on knowledge [11] Naeve, A., Nilsson, M., Palmér, M., & Paulsson, F. (2005). artefacts and in different contexts in order to unobtrusively Contributions to a public e-learning platform: leverage user interest [32]. infrastructure; architecture; frameworks; tools. International Journal of Learning Technologies , 1 (3). 3. CONCLUSION [12] Van Harmelan, M. (2006). Personal Learning This paper discussed four main factors deemed crucial for Environments. Proceedings of the 6th IEEE international developing PLE that are fit, ready for appropriation, and capable conference on Advanced learning technologies (pp. 815- of evolving over time: the adoption of social media paradigms, 816). Washington D.C.: IEEE Computer Society. the flexible representation of interaction and learning contexts [13] Grudin, J. (1988). Why CSCW applications fail: problems (including those explicitly defined by learners and those implied in the design and evaluation of organizational interfaces. from their actions), the incorporation of elastic community and Proceedings of the 1988 ACM conference on Computer- content management features encouraging the spontaneous supported cooperative work (pp. 85-93). New York: ACM. appropriation of knowledge objects, and finally the delivery of [14] Garrett, J. J. (2005 2-18). Ajax: A New Approach to Web personalized and contextualized recommendation services. We Applications. Retrieved 2010 20-06 from are currently working on Graaasp5, an online PLE that builds on AdaptivePath.com: the four PLE elements discussed in this paper. http://www.adaptivepath.com/ideas/essays/archives/000385 .php 4. REFERENCES [1] Government of Ireland Stationery Office. (2000). Learning [15] Liccardi, I., Ounnas, A., Pau, R., Massey, E., Kinnunen, P., for Life: White Paper on Adult Education. Dublin. Lewthwaite, S., et al. (2007). The role of social networks in students' learning experiences. ACM SIGCSE Bulletin , 39 [2] Aspin, D., & Chapman, J. (2007). Lifelong Learning: (4), 224-237. Concepts and Conceptions. In D. Aspin (Ed.), Philosophical Perspectives on Lifelong Learning. [16] Siemens, G. (2004 12-December). Connectivism, A Dordrecht: Springer. Learning Theory for the Digital Age. Retrieved 2010 30- June from elearn space: [3] Griffin, C. (1999). Lifelong Learning and Social http://www.elearnspace.org/Articles/connectivism.htm Democracy. International Journal of Lifelong Education , 18 (5), 329-342. [17] El Helou, S., Li, N., & Gillet, D. (2010). The 3A Interaction Model: Towards Bridging the Gap between [4] Colardyn, D., & Bjornavold, I. (2004). Validation of Formal and Informal Learning. International Conference formal, non-formal and informal learning: policy and on Advances in Computer-Human Interaction (pp. 179- practices in EU Member States. European Journal of 184). Washington D.C.: IEEE Computer Society. Education , 39 (1), 70-88. [18] Farrell, S., Lau, T., Nusser, S., Wilcox, E., & Muller, M. [5] Cross, J. (2006). Informal Learning: Rediscovering the (2007). Socially augmenting employee profiles with Natural Pathways that Inspire Innovation and Performance. people-tagging. The 20th annual ACM symposium on User New Jersey: John Wiley and Sons. interface software and technology (pp. 91-100). New York: ACM. [19] Körner, C., Benz, D., Hotho, A., Strohmaier, M., Stumme, 5 http://graaasp.epfl.ch G. (2010). Stop thinking, start tagging: tag semantics 32 emerge from collaborative verbosity. Proceedings of the 19th international conference on World Wide Web. [20] Ackerman, M. (2001). The Intellectual Challenge of CSCW: The Gap Between Social Requirements and Technical Feasibility. In J. Carrol (Ed.), HCI in the New Millennium. New York: Addison-Wesley. [21] Twidale, M. B., Wang, X. C., & Hinn, D. M. (2005). CSC: computer supported collaborative work, learning, and play. Proceedings of the 2005 conference on Computer support for collaborative learning (pp. 687-696). International Society of the Learning Sciences. [22] Wenger, E. (1998). Communities of practice: learning, meaning, identity. Cambridge, UK: Cambridge University Press. [23] LaContora, J., & Mendonca, D. (2003). Communities of practice as learning and performance support systems. Information Technology: Research and Education, 2003 (pp. 395 - 398). Washington D.C.: IEEE Computer Society. [24] Stanoevska-Slabeva, K., & Schmid, B. (2000 4-January). A generic architecture of community supporting platforms based on the concept of media. Proceedings of the 33rd Hawaii international conference on System sciences- Volume 1(p. 1014). Washington D.C.:IEEE Computer Society. [25] Dourish, P. (1992). Applying Reflection to CSCW Design. Position paper presented at Workshop on "Reflection and Metalevel Architecture" European Conference on Object- Oriented Programming. Utrecht, Netherlands. [26] Rocha, L. M. (1998). Selected Self-Organization And the Semiotics of Evolutionary Systems. In S. Salthe, G. G. Van de Vijver, & M. Delpos (Eds.), Evolutionary Systems: The Biological and Epistemological Perspectives on Selection and Self- Organization (pp. 341-358). Kluwer Academic Publishers. [27] Downes, S. (2007). Learning Networks in Practice. (D. Ley, Ed.) Emerging Technologies for Learning , 2, 19-28. [28] Pilgrim, M. (2002 18-December). What is RSS? Retrieved 2010 30-June from XML: http://www.xml.com/pub/a/2002/12/18/dive-into-xml.html [29] Brusilovsky, P., & Henze, N. (2007). Open Corpus Adaptive Educational Hypermedia. The adaptive web: methods and strategies of web personalization , 671-696. [30] Koper, R., & Tattersall, C. (2004). New directions for lifelong learning using network technologies. British Journal of Educational Technology , 35 (6), 689-700. [31] Tang, T., & McCalla, G. (2009). A Multidimensional Paper Recommender: Experiments and Evaluations. Internet Computing, IEEE , 13 (4), 34-41. [32] El Helou, S., Gillet, D., & Salzmann, C. (2010). The 3A Ranking System: Contextual, Personalized & Simultaneous Recommendation of Actors, Activities & Assets . Journal of Universal Computer Science, on Context-aware Recommender Systems, In press 33