Intelligent Adaptation and Personalization Techniques in Computer-Supported Collaborative Learning

Portable and interactive technologies are changing the nature of collaborative learning practices and open up new possibilities for Computer Supported Collaborative Learning (CSCL). Now, activities occurring in and beyond the classroom can be combined and integrated leading to a new type of complex collaborative blended learning scenarios. However, to organize and structure these scenarios is challenging and represent a workload for practitioners, which hinder the adoption of these technology-enhanced practices. As an approach to alleviate this workload, this paper proposes a proof of concept of a technological solution to overcome the limitations detected in an analysis of an actual collaborative blended learning experiment carried out in a previous study. The solution consists on a Unit of Learning suitable to be instantiated with IMS Learning Design and complemented by a Generic Service Integration system. This chapter also discusses to which extent the proposed solution covers the limitations detected in the previous study and how useful could be for reducing the orchestration effort in future experiences.

The IMS Learning Design specification is a widely known language that allows modelling of, amongst other learning designs, collaboration scripts in e-learning. Yet, it has been criticized for a number of shortcomings and specifically its lack of support for comprehensive adaptation features. We propose concrete extensions to the specification, which address a wide range of problems and omissions. The most important areas of modifications and amendments include: explicit representation of groups and corresponding collaboration contexts, as well as of artefacts as results of joint work; flexible integration of communication and collaboration services; a revamped script organization and sequencing model; a previously missing run-time model, with support for event- and exception- handling. The above are complemented by a wide range of adaptive interventions that can affect the script’s progress at run-time, tailor it to changing circumstances, and support learners. Last but not least, sophisticated scenarios are made possible through support for non-traditional collaboration script elements: the possibility to represent human involvement in adaptation decisions, ‘transactional’ action processing, loops and branches for controlling action execution, and the declaration of re-usable action sequences and complex expressions. Further to the proposed changes, examples are provided that highlight the novel possibilities afforded by these changes for advanced collaboration scripts.

In this article we present a framework for the integration of external and independent software components into IMS-LD (Learning Design) based courses that cater for adaptivity. Our framework comprises a design specification and an implementation of adaptations in CSCL (computer-supported collaborative learning) oriented and standards based architecture. The architecture allows combining existing research on explicit representations of collaborative learning processes (i.e. learning designs) with the availability of existing and tested collaborative learning tools (e.g. a forum in a virtual learning environment (VLE), an agent, a service or even a software component that provides a specific functionality). The architecture allows controlling the learning tools either by a human or a pedagogical agent and thus enables adaptive interventions to the flow of the learning activity. A mediator component is the key element in the proposed architecture. To prove the soundness of the architecture and the flexibility of its implementation example scenarios are illustrated. In these scenarios IMS-LD based modeling and Coppercore engine are used to implement adaptations by setting IMS-LD properties according to input from an external Moodle forum tool. The whole process is mediated by an integration component provided to the teacher as a Moodle resource. Finally, we highlight what would be important issues toward integrating the adaptation pattern capabilities in IMS-LD compliant tools for collaborative learning design.

This work presents the design and preliminary evaluation data regarding two prototype tools (namely, FlexCoLab and PPR), which have been designed according to the prescriptions of the adaptation pattern perspective for promoting a flexible design of CSCL activities. An adaptation pattern is described as a pedagogically useful and well-targeted adjustment process that can be initiated by the teacher or the CSCL system, in order to foster an improved learning setting when specific conditions occur during the collaborative learning activity. Both tools aim to support teachers in the process of developing flexible designs of online collaborative activities by reusing and customizing adaptation patterns, according to the requirements of a particular learning situation. “Advance the advanced” and “Support group of novices” are two adaptation patterns implemented by FlexCo- Lab, while “Advance the advanced”, “Heterogeneity group formation based on prior domain knowledge”, “Support groups of novices” and “Lack of confidence” are the adaptation patterns supported by PPR. In the chapter we present the theoretical background of adaptation patterns, the design specifications of the two systems and student evaluation data from implementing an in-school collaborative learning activity supported by PPR.

This paper presents an innovative description and an initial implementation of the “Students Team Achievement Divisions (STAD)” collaboration method (Slavin, 1978), in the form of an online adaptive collaborative design-pattern that has been constructed taking into account adaptation techniques, within the context of an open-source learning design-based environments such as the LAMS system (Dalziel, 2003). This method is described with special reference to the learning of essential aspects of an Information System. The innovative description of the aforementioned collaborative method within the LAMS system is based on the fact that: (a) the tasks assigned to the groups consist of investigation of real world scenarios, and not merely the study of learning material as is usually proposed, (b) adaptive techniques are integrated with the method and (c) for the design of the collaborative learning activity, an intuitive learning design tool such as the LAMS system is used. A research study was also conducted aiming the development of an empirical model to allow the implementation of the aforementioned adaptive STAD collaborative method within the context of an IT work place, namely; the Legal Council of the Hellenic State. In fact, the data gathered from this study were used to build the initial learning profile of the user –that is needed for the implementation of Phase 2 of the previously mentioned adaptive STAD collaborative method- so that to be able to provide him/her personalized training, monitoring, scaffolding and evaluation.

Communication is an integral part of any Computer Supported Collaborative Learning (CSCL) research approach. Furthermore the issues of Adaptation and Interaction Analysis have been intensively researched during the past years, under the scope of web-based educational approaches. Both research fields seem to share similar or complementary techniques, aims and outcomes. This chapter attempts to examine at what extend these two fields are complementary to one another or whether they can converge (or diverge) in the future, focusing on Computer Mediated Communication tools, especially Asynchronous Discussion Platforms. The existing work on applying IA methods in communication-based CSCL approaches is examined and correlated with the main constituents of the research on adaptive systems. Issues related to flexibility, adaptability and interoperability are also discussed upon, in an attempt to distinguish the future trends of the IA research field and its relation to Adaptation, outlining their conceptual similarities and examining the possibilities of developmental interconnections among the two fields.

Teachers often use flow design patterns of particular learning strategies in order to define the type of the learning and supportive tasks, their duration, their orchestration as well as the use of learning objects, tool and services needed to support the execution of these tasks. It is quite easy for teachers (even novice learning designers) to create a learning script by applying a flow design pattern. However, they often need to mix and match learning strategies in order to create customized learning scripts that are more appropriate to the learners’ preferences and the learning context in general. This task is even more challenging when such adaptations need to be made on-the-fly, i.e. during the learning process and in response to the learner’ online behavior. The aim of this paper is to discuss how the learners’ interaction data that is collected during an online learning process and analyzed using interaction analysis indicators can be used by a teacher to alter the learning script on the fly. It is shown that with the aid of a learning interaction analysis tool, which is called CoSyLMSAnalytics, a teacher can modify a learning script that is based on a typical Think-Pair-Share strategy in order to offer scaffolds to the learners during the learning process. Also, an example will be shown that will depict how a learning designer can create a variation of a learning scenario that is based on a typical Think-Pair-Share strategy by using the tool referred above.

Current studies in the field of assistive design highlight the aim to develop intelligent and interactive services capable of maintaining and augmenting support during broader conditions of decline in personal and infra structural resources (Bos, 2007). These systems are frequently challenged to produce measureable improvements in user learning and function while also responding to the attendant challenges of coherency, interoperability and usability (Manning and Benton, 2010). Such systems are frequently intended to build and transfer augmented coping capacity via the interaction of system- use and user-behaviour, essentially offering a personal learning space and functional resilience. The operating environment for many such systems can be found, for example within both the formal and informal care services. Challenges to the provision of care and the maintenance of personal choice continue to escalate, as does the pressure on systems designed to meet ever changing user needs and aspirations (Benton and Manning, 2008). In the context of health care, well documented resource and demographic challenges (Symonds et al, 2007, Manning et al, 2008), will escalate the need for system utility as a core source of user autonomy. Generally any interface between a system and user will be characterised by redundancy, a deficient use resulting from issues concerning misalignment between; accessibility, learning support and system functionality. Essential user flexibility, acceptance and acquisition of system based protocols will be influenced by the degree of ‘personalisation’ achieved by the system with a related impact on use (Rigby, 2010). In this context personalisation relates to the degree to which users’ perceptual expectations and ‘affordances’ are aligned with system design and feedback. It is suggested that individual differences of; cognitive style, techno-acceptance and personality may be evaluated to form a Behavioural Prototype©, an evaluation of users’ likely interactive expectations and behaviour with the system. This profile would provide a behavioural link between system design and a broader vision of user need, capacity and aspiration.

This chapter presents the design of an intelligent agent based system that aims to support teachers in supervising and evaluating learners and activities of lessons in the Learning Activity Management System (LAMS). A monitoring agent has been designed to collect and aggregate information from LAMS database, related to the participation of learners in lesson’s activities, at time intervals which indicates the teacher through a scheduler agent. A user notification agent diagnoses conflicts in the learning and collaborative processes and issues alert and awareness messages to the teacher, learners and groups of learners via rules that are based on a teacher defined participation model. The system, based on this model, generates also evaluation reports of learners and activities of the lesson, in order to assist the teacher to intervene effectively in the course of a lesson.

Electronic asynchronous discussion fora are increasingly becoming part of the distance education process and are a dynamically evolving field which needs to be constantly updated and redefined. This chapter presents a system development approach for automated interpretation of discussion threads’ messages in asynchronous distance education fora by using the content category as unit of analysis. This system inputs discussion threads from distance education fora and outputs specific strings representing the messages of these threads according to concrete modelling based in a formal language.

Massively Multiplayer Online Games (MMOGs) are rich in goal-oriented activities and collaborative and social interactions, both essential for learning the game and progressing. In this chapter we employ a theoretical framework for linking learning and collaborative learning principles with MMOGs and investigate, through an exploratory and qualitative approach, features of the tasks, groups, and player interactions that may support the emergence of collaborative interactions and learning. The critical role of both the design of the environment and of the community of players is highlighted and it is concluded that their balanced inter-connection is critical for the emergence of effective collaborative interactions. Implications on further research are also discussed.

Interdisciplinarity in the Computer Supported Collaborative Learning (CSCL) research field involves the application of several methodological approaches towards analysis that range from deep-level qualitative analyses of small interaction-rich episodes of collaboration, to quantitative measures of suitably categorized events of interaction used as indicators of the success of collaboration in some of its facets. This article adopts an alternative approach to CSCL analysis that aims at taking advantage of some desired properties of each of these diverse methodological trends, involving the use of a rating scheme for the assessment of collaboration quality. After defining a set of dimensions that cover the most important aspects of collaboration, it employs appropriately trained human raters basing their assessments on substantial aspects of collaboration that are not easily formalisable. The activities studied here regard 228 collaborating dyads, working synchronously on a problem-solving task. Based on this large dataset, relations between dimensions of collaboration quality are unraveled on empirical grounds, by elaborating ratings statistically using a multidimensional scaling technique.

A community of practice is a group of people who share common concerns, problems or passions for a domain and who deepen their knowledge and expertise through interaction and collaboration on an ongoing basis. More and more people, groups and organizations are looking to develop Internet-mediated communities of practice in order to realize specific goals on informal learning and professional development. Harnessing the perceived values of communities across educational sector requires well-designed settings and procedures to achieve a sustainable level of functionality, communality, collaboration and knowledge sharing. Because current research supports the notion that there is not a systematic theory or a blueprint for design of online communities, this work aims to define a basic typology of various critical elements for successful and sustainable Internet-mediated communities of practice, via a meta-analysis and critical synthesis of related literature.