Contextualised Media for Learning Tim de Jong, Marcus Specht and Rob Koper Educational Technology Expertise Centre, Open University of the Netherlands, Valkenburgerweg 177, 6419 AT, Heerlen, The Netherlands tim.dejong@ou.nl, marcus.specht@ou.nl, rob.koper@ou.nl Abstract: This paper motivates contextualized media for learning and starts with a reference model for describing and analysing contextualized media applications for learning. Based on a more detailed analysis of the context dimension and contextual parameters taken into account when supporting mobile and ubiquitous learning a technical framework and its components are presented. The framework allows to model and implement complex educational scenarios for contextualized learning and integrates different components and layers necessary. As an example application “ContextBlogger” is described which implements contextualized blogging support based on APIs of standard blogging systems and different context sensors as GPS and Semacode. The paper concludes with an outlook on possible chances and challenges for future research. Keywords: Contextualised learning, m-learning, language learning, mobile social software, ubiquitous computing

Introduction Contextualized and ubiquitous learning are relatively new research areas that combine the latest developments in ubiquitous and context aware computing with pedagogical approaches relevant to structure more situated and context aware learning support. Searching for different theoretical underpinnings of mobile and contextualized learning (Naismith, Lonsdale, Vavoula, & Sharples, 2004) have identified the relations between existing educational paradigms and new classes of mobile applications for education. Furthermore best practices of mobile learning applications have been identified and discussed in focused workshops (Stone, Alsop, Briggs, & Tompsett, 2002; Tatar, Roschelle, Vahey, & Peunel, 2002). Especially in the area of educational field trips (Equator Project, 2003; RAFT, 2003) in the last years innovative approaches for intuitive usage of contextualized media have been developed. Situated Learning and communities of practice (Wenger & Lave, 1991) stress the importance of embedding the learner in an authentic context and in a peer community to reflect about her learning. Furthermore, approaches like reflection in action and reflection about action describe the relevance of the context for enabling learning and self reflection (D. A. Schön, 1983; D.A. Schön, 1987). In our current research in the TENCompetence we investigate the role of supporting informal learning activities and integrating them with formal and lifelong learning approaches in learning networks (Rob Koper, 2005). From our point of view, the role of ubiquitous support for learning activities in learning networks is essential for embedding learning into every-day living and to support situated and informal learning in learning networks. Additionally social software for learning like educational blogs have recently become a popular way of collecting personal information and learning experiences (Oravec, 2002) and combine this with the reflection in a community – the social context. Mobile blogging provides an instant way of accessing and collecting personal memories. The added value for personal reflection or community building have been researched in undergraduate and higher education (RAMBLE-Project, 2006; Marcus Specht & Kravcik, 2006). In that sense blogs are simple tools for supporting long-term informal learning processes (Trafford, 2005) embedded in a community and fostering reflection. In context-aware computing a variety of notions of context have been discussed and automatic possibilities for context detection, context matching and sensor tagging have been researched (Abowd & Mynatt, 2000; Dey & Abowd, 2000). Context-aware computing together with ubiquitous and pervasive techniques can result in systems that adapt to user’s identity, preferences, location, environment and time (Gross & Specht, 2001; Marcus Specht & Kravcik, 2006; A. Zimmermann, Specht, & Lorenz, 2005). For building contextualized learning support on the one hand an infrastructure for contextualization is needed. This builds on research works in the area of context aware systems (A. Zimmermann, Lorenz, & Specht, 2005). On the other hand methods for analyzing

and designing context specific tools for learning support are necessary (M. Specht, 2007). Third from a human computer interaction perspective (Terrenghi, Specht, & Moritz, 2004) new methods of interacting with ubiquitous and contextualized media and learning experiences need to be researched. In the following paper we will define a framework based on our experiences building contextualized media applications for giving contextualized support to learners. The paper is structured as follows. The second section provides a reference model for contextualised media in learning and a analysis of the contextual parameters. Then, based on the theory a technical framework is described in section three. Section four then describes an example application of contextualised blogging for learning. Finally, section five provides a summary of the paper and conclusions.

Why Contextualised Media? Situated learning as introduced by Lave and Wenger (Wenger & Lave, 1991) states the importance of knowledge acquisition in a cultural context and the integration in a community of practice. Learning in this sense must not only be planned structured by a curriculum but also by the tasks and learning situations and the interaction with the social environment of the learner. This is often contrasted with the classroom-based learning where most knowledge is out of context and presented de-contextualized. On the one hand the process of contextualization and de-contextualization might be important for abstraction and generalization of knowledge on the other hand in the sense of cognitive apprenticeship (Collins, Brown, & Newman, 1989) it is reasonable to guide the learner towards appropriate levels and context of knowledge coming from an authentic learning situation. From a constructivist point of view not only knowledge is always contextualized and but also the construction of knowledge, e.g. learning is always situated within its application and the community of practice (Mandl, Gruber, & Renkl, 1995). Stein defines four central elements of situated learning where the content emphasizes higher order thinking rather than acquisition of facts, the context for embedding the learning process in the social, psychological, and material environment in which the learner is situated, the community of practice that enables reflection and knowledge construction, and the participation in a process of reflecting, interpreting and negotiating meaning (Stein, 1998). From the perspective of situated learning several requirements for new learning tools can be stated like: use authentic problems, allow multiple perspectives, enable learning with peers and social interaction within communities, enable active construction and reflection about knowledge. A shift towards a new tradition of online learning is described by Herrington et. al. (Herrington, Oliver, Herrington, & Sparrow, 2002). Moreover the idea of situated learning is also closely related to the ideas of “blended learning” and “learning on demand” especially in educational systems for adults and at the workplace (Oppermann & Specht, 2006). An important point that is not taken into account by a lot of new approaches for delivering learning on demand is the aspect that the need (demand) for knowledge and learning arises in a working context with the motivation for solving specific problems or understanding problem situations. This notion of “learning on demand” in the workplace exemplifies the potential of contextualized learning in the workplace. Learners that identify a problem in a certain working situation are highly motivated for learning and acquiring knowledge for problem solving. They have a complex problem situation as a demand, which can be used for delivering learning content adapted to their situation. Furthermore not only the delivery of content into a certain context or practice is needed but also interaction facilities must be provided which allow an appropriate interaction and cooperation with educational systems. The contextualization of the learning on demand can not only be seen from the point of view of an actual problem or learning situation but also in a longer lasting process of learning activities that are integrated. Different learning activities are combined in blended learning approaches where the preparation for a task, updates on base knowledge, then the application in an actual working situation and the documentation of problem solutions and the reflection about one’s activities evaluates that process. Latest research stresses also two other dimensions of embedding learning support into everyday life: first integration from a lifelong learning perspective and therefore second also in a community of practice. Latest research into lifelong learning integrates informal and formal learning approaches and supports access to knowledge resources, learning activities, competence development, and learning communities from a variety of clients build on service oriented architectures (R. Koper & Specht, 2006).

An extended reference model for contextualised learning In (deJong, Specht, & Koper, to appear) the authors present a review of current systems for mobile contextualized learning support and a reference model that is used to classify the current research. The reference model is used to identify blind spots in the current research and to discuss new solutions for contextualized learning support. Table 1 shows the dimensions of the reference model with the possible values for each dimension.

Table 1 A reference model for mobile social software Content Context Information

flow Purpose Pedagogical

model

Annotations

Documents

Messages

Notifications

Individuality Context

Time Context

Locations Context

Environment or Activity Context

Relations context

One-to-one

One-to-many

Many-to-one

Many-to-many

sharing content and knowledge

facilitate discussion and brainstorming

social awareness

guide communication

engagement and immersion

behaviourist

cognitive

constructivist

social constructivist

The reference model describes the type of content that is used in contextualized learning tools, the context parameters taken into account for adaptation, the information flow, and on a higher level the main purpose and the underpinning pedagogical model.

• The content dimension describes the artefacts exchanged and shared by users, in an analysis of the literature the main types of artefacts found were annotations, documents, messages, and notifications.

• The context dimension describes the context parameters taken into account for learning support. The main context dimensions identified based on an operational definition of context (Andreas Zimmermann, Lorenz, & Oppermann, 2007).

• The information flow, classifies applications according to the number of entities in the systems information flow and the information distribution. The classification also integrates mobile social software applications in contextualized learning support.

• The purpose, describes applications according to the goals and methods of the system for enabling learning.

• The pedagogical paradigms and instructional models describe the main paradigm leading the design of contextualized media and the integration of media in real world contexts.

By combining different values of those underlying dimension various forms of contextualized software can be created. For instance, a classical location-based information system combines the access to documents with a locations context, a one-to-many information flow to guide communication and strengthen engagement and immersion. Moreover, the combination of shared information in a specific location context can be used to support a learning process based on a social constructivist pedagogy. Other combinations lead to other kinds of systems which have already been explored extensively in (deJong, Specht, & Koper, to appear). The dimensions of the reference model are not entirely independent. The two last dimensions in table 1, purpose and pedagogical model, can be described using the three other dimensions of content, context and information flow. Hence, the reference model consists of three lower level dimensions, with two higher level dimensions building on top of them. A system that has sharing content and knowledge as it main purpose can be described by using documents from the content dimension, relations context, and a many-to-many information flow.

Likewise, the pedagogical model used can also be described using the three lower level dimensions. In an educational setting typically learning content is exchanged between a source and a recipient. The information flow dimension from the reference model dictates this exchange of content because it specifies how many sources and how many recipients take part in the educational setting. Finally, learning takes place in a specific real-world setting which can be described using the context parameters of the reference model. A traditional classroom setting can now be described as follows. The content is exchanged between a teacher and a number of students; especially documents and messages (speech) are exchanged between them. The information flow ranges from one-to-many, in which the teacher addresses the students, one-to-one, in which one student asks a question and, many-to-one, in case the students submit their tests to the teacher for correction. Moreover, the context dimension can be used to describe the individual context of each of the participants separately. A shared context can be derived from overlapping individual context information, that describes the knowledge of the group (Andreas Zimmermann, Lorenz, & Oppermann, 2007). How the overlap between individual and group contexts is used is a main design criteria of contextualized groupware applications and dependent on the purpose that a system aims for. In most approaches overlaps between individual contexts are used for social awareness and facilitation of cooperation in learning settings.

Which Context Parameters? The context dimension describes the kind of contextual information that is used in the system; contextual information can be used to describe or derive information about the user (describing for example the learner’s personal preferences), information about the environment, (describing the learner’s physical environment) or, information about the social context of the learner (describing the social relationships a learner is involved in and the social networks the learner is part of). In community systems the context describes where a user is located within the community (boundary, centre) and the number and kind of relations this user has to other users within the community. In context-aware computing a variety of notions and interpretations has developed over the years. Zimmermann et. al distinguish between definitions by synonym or definitions by example which mainly name and describe certain context parameters as location, identity, time, temperature, noise, as well as beliefs, desires, and commitments and intentions (Andreas Zimmermann, Lorenz, & Oppermann, 2007). Furthermore they introduce an operational definition of context describing following main categories of context information:

• Individuality Context, includes information about objects and users in the real world as well as information about groups and the attributes or properties the members have in common.

• Time Context, basically this dimension ranges from simple points in time to ranges, intervals and a complete history of entities.

• Locations Context, are divided into quantitative and qualitative location models, which allow to work with absolute and relative positions.

• Activity Context, reflects the entities goals, tasks, and actions. • Relations Context, captures the relation an entity has establish to other entities, and describes

social, functional, and compositional relationships. In the literature for mobile and contextualized learning support different examples can be found for

utilizing those context parameters to enable learning. Several projects have looked at how to use and enrich contents with contextual metadata (Equator Project, 2003; Marcus Specht & Kravcik, 2006). Mostly interesting new approaches in context-aware systems see the main strength in combining different context parameters for user support. Even more new approaches tend to combine different forms of metadata about learning objects and media to allow for flexible and contextualized learning support. In the MACE project the combination of various types of content, usage, social and contextual metadata enables users to develop multiple perspectives and navigation paths that effectively lead to experience multiplication for the learner himself (Stefaner et al., 2007). Identity context is also often combined with other forms of context. One specific example of such a combination is given by (Ogata & Yano, 2004b) who present CLUE, a system for learning English in real-world situations. CLUE uses (1) a learner profile to adapt the learning content to the learner’s interest and (2) location information to link objects/locations to suitable English expressions, i.e. appropriate learning content. Likewise, (Bo,

2002) combines a user profile and user position, to facilitate personalised and location-based information delivery. AwarePhone (Bardram & Hansen, 2004) uses several context-parameters at the same time. First of all, location is used to locate fellow employees within the hospital. Second, a calendar artefact is used to capture and share time context and also indicate the activity of a user at a certain moment. The activity is furthermore given by a shared status message. The combination of these three context parameters lead to what the writers call “context-mediated social awareness”.

Environmental context information is used in several systems, most notably QueryLens (Konomi, 2002) which focuses on information sharing using smart objects. Moreover, the TANGO system presented by (Ogata & Yano, 2004a) and the Musex system (Yatani, Sugimoto, & Kusunoki, 2004) detect objects in the vicinity by using RFID tags. Moop (Mattila & Fordel, 2005) couples a GPS location to observations/information gathered in the field for later analysis in the classroom. Wallop (Farnham, Kelly, Portnoy, & Schwartz, 2004) allows its users to discover social relationships and provides social awareness by storing and analysing social context information; to derive the social context communication patterns system interactions and co-occurrence information were analysed.

A technical framework for contextualised media The technical framework will be based on the reference model described earlier. The reference model consisted of two levels; the lower level describes what the contextualise media infrastructure manipulates (content, context) and moreover where the data is manipulated and between which subsystems and actors the data flows (information flow). Conversely, the higher level describes how to group the lower level concepts to come to certain functionality or to organise an educational process. To support the contextual media mentioned in the previous section a flexible technical framework is necessary. One part of our technical framework for contextualised media is based on already existing infrastructures for context management. Zimmermann, Specht and Lorenz (A. Zimmermann, Specht, & Lorenz, 2005) suggest a standard architecture for context management that semantically enriches contextual data step by step in successive layers. On the other hand through our work in implementing different contextualized media solutions it became obvious that a flexible approach for handling content and combining content and context in delivery and content creation in applications is essential. Therefore beside the context management part we introduce and define an independent media handling part of the framework that allows us to handle different types of contents on an abstract level.

Figure 1: The basic contextualized media framework its layers and entities.

The whole framework is realized as a service-oriented architecture, consisting of a server and several clients that provide the sensors and actuators (Rehrl, Bortenschlager, Reich, Rieser, & Westenthaler, 2004). The information flow between the separate subsystems will be handled by an event-based architecture. The functional modules part of a layer can post events to an event bus and subscribe to events posted by others. When an event of interest has been generated, the specific listener is notified and can react to the event. A sensor for example can post a sensor update event with new sensor values as an event on the bus, which will be picked up by other modules listening to sensor updates. In general the context metadata system stores contextual metadata that can be used to identify a certain context. The information available from the client sensors is used to find the appropriate contextual metadata and identify a physical object or location. The contextual metadata subsystem can query and retrieve the content related to a certain context. To filter the content according to the context, the subsystem stores relations between context tags and content objects. The context subsystem consists of four layers in which the enrichment of context takes place. On the sensor layer classical GUI components and the user’s interactions with them can be used to record information about the user’s context or data that was created by users. On the semantic layer we model different basic types of artefacts exchanged and used by applications and the entities in the real world that can be connected to them. Furthermore aggregation of sensor values takes place for instance, by combining location and time sensor data the aggregation layer can derive speed information of a subject. On the control layer we describe educational scenarios in IMS LD where we can combine properties of physical world objects or user properties and integrate them in IMS-LD descriptions based on the entities defined in the semantic layer. Basically two types of components are found on this layer:

• Content-Context Modelling: rules determining what content should be delivered based on the current context information available, or a subset of that information. For example, a rule can be create that sparks a notification when near a specific location.

• Educational Process Modelling: pedagogical strategies and learning processes can be modelled by using IMS Learning Design (R. Koper, Olivier, & Anderson, 2003). Context information can be used to drive the modelled learning process into a specific direction. In this way, the modelling of contextualised learning processes with IMS-LD would be possible.

On the indicator and actuator layer we define input and output devices and metatag them with contextual metadata, on the one hand this allows to convert content and adapt them to the properties of the output devices or even to select appropriate output devices in the learner’s environment, if for example the noise level is too high for people to hear an audio feedback, visual feedback should be chosen. On the media level the semantic layer handles all content created by the users and provides a way to query and retrieve that content. Via abstract entities the subsystem can handle all sorts of content, a range of multimedia content, and therefore should provide corresponding methods of retrieval (for instance, media download or streaming). Each information object is stored with a unique identifier associated to it, for later retrieval. In addition, the content subsystem provides different levels of content aggregation; for example by aggregating content into categories. Media processing and adaptation between different formats is done on the indicator/actuator layer. Each of the content repositories should be accessible via a standardised interface. The decision about where to store certain injected content takes place in this layer. Moreover, the content repositories will be queried by either queries from one of the clients or from the context subsystem. To facilitate client specific content delivery, this layer has to know about the displaying capabilities of each device or the kind of actuator that will be used to display the content.

Applications of the technical framework The extension of the reference model and the technical framework described in the previous sections allow us to (1) model different educational applications based on three dimensions of content, context and information flow and (2) implement these educational applications in a standardised way, with minimised effort. As one example we will describe contextBlogger (De Jong, Al Takrouri, Specht, & Koper, 2007) which from our view demonstrates the possibilities of the framework described above. Contextual blogging combines social software, a weblog, with information about the context of a learner. The information in the weblog can be accessed using a mobile device, and the content can be filtered through the application of search filters based on context information. The search filters for the contextual blogging application retrieve the content either related to a specific real-world object or to a specific user location. Furthermore, the learner can also choose to create his/her own content and relate it to a real-world objects or locations. Therefore, the use of the contextual blogging application provides a basis for an investigation of the usage of physical artefacts in learning. On the one hand the combination with a physical object could provide the basis for learning, on the other, shared objects could be used to build communities of practice and couple the creation of learning networks to physical objects. Through applying different context filters in combination with the creation or retrieval of weblog content, we expect to achieve different educational effects:

• Multiple perspectives on real-world objects: by viewing the object’s history, a certain category of blog entries, or using other filters people benefit through an indirect learning process (Efimova & Fiedler, 2004; Walker, 2005).

• Community-generated content connected to relevant real-world objects and locations: an example for the effect and importance of self-generated contents in a learning community is presented in (Eva Brandt, Björgvinsson, Hillgren, Bergqvist, & Emilson, 2002; Eva Brandt & Hillgren, 2003) about learning to operate medical devices.

• Community interaction and the creation of communities of interest around certain objects and locations, supporting contextualised learning.

• Different views about objects, based on personal preferences. Real-world objects can also be linked electronically to create relations between those objects and to create a so-called “internet of objects” (Mattern, 2004).

• Increase motivation through active learning, by actively involving the learner in the learning process, the learner involvement and motivation is increased. This as opposed to passive learning in a formal classroom setting.

To achieve these educational effects the underlying concepts of a system for contextualised blogging and the relations between them should be analysed. For instance, to create multiple perspectives on real-world objects and locations, a user should be able to interact with a physical object and should be

able to retrieve content linked to that physical object. By using shared real-world objects, multiple users can interact with them, and create information objects related to them or view, rate and comment the content added by other people (community-generated content). In that way, a community of users can evolve around these shared objects and the community interaction leads to different opinions and perspectives about these objects. The multitude of perspectives about a shared object, can lead to either a discussion between users with different opinions or leads to reflection about a situation by the learner; either by looking at the opinions of other users, or by adding content and reading it back later, as an opportunity to reflect back on what happened before (D. A. Schön, 1983; D.A. Schön, 1987). To prevent the user from being overwhelmed by the amount of information available in a community contextualised search filters are used that only display the relevant information for a certain situation or context. By combining these educational effects the system addresses the lifelong learner, by providing several opportunities for the self-centred learner or a community of these learners to structure the learning process. Also the system relies on the implicit assumption of lifelong learning that responsibility for the creation and structuring of learning content resides with the self-directed learner himself (Rob Koper & Tattersall, 2004). Applying the technical framework described above allows us describe different contextual blogging applications and how the fulfil the pedagogical aims described in the reference model.

Figure 2: Educational applications of contextualized blogging

By giving a user a single blog mainly a (1) personal learning history and reflection about this history is fostered. Secondly by combining a context tag and therefore physical objects with a user blog a (2) contextualized learning history and reflection about the learner’s activity and about physical objects and activities are supported. Furthermore users can reflect about a single instance of an object or about a abstract class of objects by combining objects with single blog entries (5) or blog categories (4). The detailed implementation of the contextualized blogging system is described in (De Jong, Al Takrouri, Specht, & Koper, 2007).

Conclusions

In this paper, a reference model for mobile social software for learning was described. The reference model consisted of five dimensions, being: purpose, content, context filter, information flow and pedagogical model. Based on that model mainly different context filters used for contextualized media where analysed and a technical framework was presented that allows to combine content, context and different application logics for contextualized media applications. As one example the context blogger application was described that allows learners to exchange and share contextualized media. Dependent on how the information flow and the media filtering are defined it has been shown that different educational effects and objectives can be achieved.

For future work we will empirically evaluate the effects of contextualized social media in different learning applications and validate our technical framework with different media types and context sensoring and tagging methods to be integrated. Finally, the use of more complex forms of context information derived from the five basic context parameters individuality, locations, time, environment/activity forms a challenge that future research should, we feel, certainly address. The use of social or relations context, in addition to the earlier mentioned context parameters, could provide us with contextualized learning solutions that not only address a specific individual, but also addresses the social environment and relationships for improved learning. Only a combined approach taking into account multiple aspects of a learner’s environment would make learning truly ubiquitous.

References Abowd, G. D., & Mynatt, E. D. (2000). Charting Past, Present, and Future Research in Ubiquitous

Computing. ACM Transactions on Computer-Human Interaction, 7(1), 29-58. Bardram, J. E., & Hansen, T. R. (2004). The AWARE architecture: supporting context-mediated social

awareness in mobile cooperation. Paper presented at the ACM conference on Computer supported cooperative work.

Bo, G. (2002). MOBILEARN: Project final report [Electronic Version]. Retrieved 1st July from http://www.mobilearn.org/download/results/public_deliverables/MOBIlearn_Final%20Report.zip.

Brandt, E., Björgvinsson, E., Hillgren, P.-A., Bergqvist, V., & Emilson, M. (2002, October 19-23, 2002). PDA´s, Barcodes and Video-films for Continuous Learning at an Intensive Care Unit. Paper presented at the NordiCHI.

Brandt, E., & Hillgren, P.-A. (2003, 19-20 May). Self-produced video to augment peer-to-peer learning. Paper presented at the MLEARN 2003, London, United Kingdom.

Collins, A., Brown, J. S., & Newman, S. E. (1989). Cognitive Apprenticeship: Teaching the Craft of Reading, Writing, and Mathematics. In L. B. Resnick (Ed.), Knowing, Learning and Instruction (pp. 453 - 494). Hillsdale N.J: Lawrence Erlbaum Associates.

De Jong, T., Al Takrouri, B., Specht, M., & Koper, R. (2007). Campus Memories: Learning with Contextualised Blogging Paper presented at the Conference Name|. Retrieved Access Date|. from URL|.

deJong, T., Specht, M., & Koper, R. (to appear). A Reference Model for Mobile Social Software for Learning. International Journal of Continuing Engineering Education and Life-Long Learning (IJCEELL).

Dey, A. K., & Abowd, G. D. (2000). Towards a Better Understanding of Context and Context-Awareness. Paper presented at the Conference Name|. Retrieved Access Date|. from URL|.

Efimova, L., & Fiedler, S. (2004, 24-26 March 2006). LEARNING WEBS: LEARNING IN WEBLOG NETWORKS. Paper presented at the Web-based communities, Lisbon, Portugal.

Equator Project. (2003). The EQUATOR Interdisciplinary Research Collaboration. 2003, from http://www.equator.ac.uk/

Farnham, S., Kelly, S. U., Portnoy, W., & Schwartz, J. L. K. (2004). Wallop: designing social software for co-located social networks. Paper presented at the 37th Annual Hawaii International Conference on System Sciences, 2004., Hawaii.

Gross, T., & Specht, M. (2001). Awareness in Context-Aware Information Systems. Paper presented at the Proceedings of the Mensch und Computer - 1. Fachübergreifende Konferenz, Bad Honnef, Germany.

Herrington, J., Oliver, R., Herrington, T., & Sparrow, H. (2002). Towards a new tradition of Online Instruction: Using Situated Learning Theory to Design web-Based Units. Paper presented at the 17th Annual ASCILITE Conference, Lismore.

Konomi, S. I. (2002, September 29 - October 1, 2002). QueryLens: Beyond ID-based information access. Paper presented at the UbiComp 2002: Ubiquitous Computing : 4th International Conference Göteborg, Sweden.

Koper, R. (2005). TENCompetence: Building the European Network for Lifelong Competence Development (Project Draft). Heerlen: Open Universiteit Nederland.

Koper, R., Olivier, B., & Anderson, T. (2003). IMS Learning Design Information Model: IMS Global Learning Consortium.

Koper, R., & Specht, M. (2006). TenCompetence: Lifelong Competence Development and Learning. In D. M.-A. Sicilia (Ed.), Competencies in Organizational E-Learning: Concepts and Tools: in press.

Koper, R., & Tattersall, C. (2004). New directions for lifelong learning using network technologies. British Journal of Educational Technology, 35(6), 689–700. Mandl, H., Gruber, H., & Renkl, A. (1995). Situiertes Lernen in multimedialen Lernumgebungen. In L.

J. Issing & P. Klimsa (Eds.), Information und Lernen mit Multimedia (pp. 167 - 178). Weinheim: Psychologie Verlags Union.

Mattern, F. (2004, November 2004). Wireless Future: Ubiquitous Computing. Paper presented at the Wireless Congress, Munich, Germany.

Mattila, P., & Fordel, T. (2005). MOOP - Using m-learning environment in primary schools. Paper presented at the Conference Name|. Retrieved Access Date|. from URL|.

Naismith, L., Lonsdale, P., Vavoula, G., & Sharples, M. (2004). Literature Review in Mobile Technologies and Learning (Literature Review No. 11): University of Birmingham.

Ogata, H., & Yano, Y. (2004a). Context-aware support for computer-supported ubiquitous learning. The 2nd IEEE International Workshop on Wireless and Mobile Technologies in Education (WMTE '04), 27-34.

Ogata, H., & Yano, Y. (2004b). Knowledge awareness map for computer-supported ubiquitous language-learning. Paper presented at the The 2nd IEEE International Workshop on Wireless and Mobile Technologies in Education (WMTE '04).

Oppermann, R., & Specht, M. (2006). Situated Learning in the Process of Work. In D. Hung & M. S. Khine (Eds.), Engaged Learning with Emerging Technologies (pp. 69-89). Berlin: Springer Verlag.

Oravec, J. (2002). Bookmarking the World: weblog applications in education. Journal of Adloescent and Adult Literacy, 45(7), 616-621.

RAFT. (2003). RAFT Project Website. 2003, from http://www.raft-project.net RAMBLE-Project. (2006). Remote Authoring of Mobile Blogs for Learning Environments. Retrieved

11-14-2006, 2006, from http://ramble.oucs.ox.ac.uk/index.html Rehrl, K., Bortenschlager, M., Reich, S., Rieser, H., & Westenthaler, R. (2004, 15-17 September

2004). Towards a Service-Oriented Architecture for Mobile Information Systems. Paper presented at the IFIP TC8 Working Conference on Mobile Information Systems (MOBIS), Oslo, Norway.

Schön, D. A. (1983). The Reflective Practitioner: How Professionals think in Action. London: Maurice Temple Smith.

Schön, D. A. (1987). Educating the Reflective Practitioner. San Francisco: Jossey-Bass. Specht, M. (2007). Contextualized Learning Support. In H. Adelsberger, Kinshuk, J. Pawlowski & D.

Sampson (Eds.), International Handbook on Information Technologies for Education and Training. Heidelberg: Springer.

Specht, M., & Kravcik, M. (2006). Authoring of Learning Objects in Context. International Journal on E-Learning, 5(1), 25-33.

Stefaner, M., Dalla Vecchia, E., Condotta, M., Wolpers, M., Specht, M., Apelt, M., et al. (2007). MACE – Enriching Architectural Learning Objects for Experience Multiplication. Paper presented at the 2nd european Conference on Technology Enhanced Learning EC-TEL.

Stein, D. (1998). Situated Learning in Adult Education. 2003, from http://ericacve.org/docs/situated195.htm

Stone, A., Alsop, G., Briggs, J., & Tompsett, C. (2002, 20-21st June, 2002). M-Learning and E-Learning: a Review of Work Undertaken by the Learning Technology Research Group, Kingston University, UK. Paper presented at the Proceedings of the Europeain Workshop on Mobile and Contextual Learning, The University of Birmingham, England.

Tatar, D., Roschelle, D., Vahey, P., & Peunel, W., R. (2002). Handhelds go to School: Lessons Learned.

Terrenghi, L., Specht, M., & Moritz, S. (2004). Design of an Interface for Technology Supported Collaborative Learning - the RAFT Approach. Paper presented at the International Conference on Entertainment Computing, Eindhoven, The Netherlands.

Trafford, P. (2005). Mobile Blogs, Personal Reflections and Learning Environments. ARIADNE, Web magazine for information professionals in archives, libraries and museums(44).

Walker, J. (2005). Weblogs: Learning in Public. On the Horizon, 13(2), 112-118. Wenger, E., & Lave, J. (1991). Situated Learning: Legitimate peripheral participation. Cambridge,

New York: Cambridge University Press.

Yatani, K., Sugimoto, M., & Kusunoki, F. (2004). Musex: a system for supporting children's collaborative learning in a museum with PDAs. Paper presented at the The 2nd IEEE International Workshop on Wireless and Mobile Technologies in Education (WMTE '04).

Zimmermann, A., Lorenz, A., & Oppermann, R. (2007). An Operational Definition of Context. Paper presented at the Context 07.

Zimmermann, A., Lorenz, A., & Specht, M. (2005). Personalization and Context-Management. User Modeling and User Adaptive Interaction (UMUAI), Special Issue on User Modeling in Ubiquitous Computing, 15(3-4), 275-302.

Zimmermann, A., Specht, M., & Lorenz, A. (2005). Personalization and Context Management. User Modeling and User-Adapted Interaction, 15, 275-302.