CSCL Workshop Challenge on Adaptation

De Wiki du LAMA (UMR 5127)
Aller à : navigation, rechercher

Competitive Challenge on Adapting Activities Modeled by CSCL Scripts

Workshop at CSCL'2009 - Monday, June 8th, 2009.

Call for "modeling solutions" closed.

Program and Organizing committee

Christine Ferraris, Laurence Vignollet, Christian Martel, Scenarios Team, Syscom, Université de Savoie, France
Andreas Harrer, Catholic University Eichstätt-Ingolstadt, Germany
Yannis Dimitriadis, GSIC, University of Valladolid, Spain,

Contact: Christine Ferraris

Theme and Goals of the workshop

Both the Learning Design community and the CSCL scripting community are working on the problem of modeling learning activities. The idea is to build a formal description of such activities (called a script, or a scenario, or a sequence, or a learning design) that a computer will be able to read and interpret in order to automatically generate the CSCL environment that will support the execution of the corresponding activity. Researchers of these communities have thus proposed and developed models, languages, methodologies and tools to support this approach.

A script is written in a given language (for example IMS-LD, LDL, the collaborative scripting language, etc.) and transformed in an activity running on an existing virtual learning environment. Thus the activity is strictly running according to what is described in the script: it cannot deviate as it is scaffolded by a computer process. The problem then is the following: how can it be adapted to events that may occur in the activity. Indeed, learning activities are continuously evolving whilst they are running. Teachers always have to adapt the situation to the learners’ reactions and to external events. Some may have been planned and the teacher knows how to react; some may be unpredictable and the teacher has to imagine a way to adapt.

The theme of this workshop is the adaptation of activities modeled by CSCL scripts according to what is happening in the activity. The issues to deal with include the following:

  • Can adaptation be taken into account in the scripting approach?
  • What are the models and languages required to do that adaptation?
  • Do the currently existing models, languages and supporting tools handle this problem?
  • Which concepts do these currently existing models, languages and supporting tools propose for that?
  • How can adaptation be handled?

These issues may be considered in both the case of foreseen events (adaptation is specified within the script) and the case of unforeseen events (there is nothing in the script to define how to react).

The adaptation problem raises another problem: the 'observation' one. The workshop will also deal with this problem issue: how can an activity be observed while it is going on? Are concepts dedicated to observation needed in the script models? What currently exists to support observation within scripting approaches? Which models? Which tools?</p>

Format of the workshop: a challenge with one benchmark scenario

The workshop is defined as a kind of competition gathering teams that will have to work on the design of a common case study (a given learning activity which will serve as a benchmark) dealing with the issues mentioned above. Two kinds of participants are expected:

  • “competitors”, who will be engaged in the challenge,
  • “attendees”, who will be able to look at the concrete competitors’ solutions, to get hands-on experience with these solutions and to discuss further requirements with the competitors.

It is open to any academic or industrial e-learning researcher, e-learning designer or e-learning practitioner in CSCL. It will allow confronting the various solutions and models, approaches and tools through modeling experiences of collaborative learning activities, considering the adaptation and observation problems, starting from a real situation to be modeled and ending at the corresponding implemented activity.

Planned activities

Pre-workshop – call for "modeling solutions"

A "challenge”, i.e. a scenario that should be modeled by the different competitors, has been issued. Potential contributors (future competitors) are expected to submit an initial text describing briefly their solution approach and the issues tackled in the solution for a short reviewing process to select competitors.

The selected competitors are expected to prepare a solution to the challenge for the workshop event.

During the workshop (one full day)

  • Presentation of the competitors’ solution and discussion during the morning session;
  • Open space in the afternoon for the participants to test the solutions and play with them. This will be followed by a concluding session, summarizing and comparing the approaches, and identifying further ToDos for the field.


The competitors' solutions will be collected into a technical report on “adaptation for CSCL scripts” and published at least in the web. We are currently considering a publication for a special issue of a journal.

Submission procedure and deadlines

Potential contributors (future competitors) are expected to submit a text describing briefly their solution approach (planned modeling technique to be used, intended CSCL tools, adaptation approach) and the issues tackled in the solution. A template is provided to help the contributors provide the expected information.

As we want competitors to have time in preparing the solution and not to be overbooked with writing tasks, the required description is voluntarily short. The final decision to be engaged as a competitor in the workshop will be taken on the basis of the description submitted.

If you want to apply as a competitor, please fill the form in and send it to Christine Ferraris no later than April 6th.

Notification for acceptance as a competitor will be sent by April the 10th (deadline for early registration for CSCL is April the 20th).

The proposed case study


A case study to be modeled as well as potential issues related to the adaptation and observation problems are proposed to potential contributors (future competitors). The issues are described below and illustrated in relation to the case study. Of course, contributors are not expected to consider all the issues and to provide solutions for all of them. Contributors are neither obliged to consider the listed issues as a restrictive list.

The same holds for the case study. It is proposed as an example which facilitates the description of the adaptation issues (as they can be easily situated within this example) and it may serve as a means to facilitate the comparison of the contributors’ proposals. Contributors are free to adapt it and even to use whatever case study they want (please indicate this in your submission).

Rough description of the case study

The activity to model is made of two parts: an individual work part (quiz part) followed by a collaborative one (discussion part). In the first one, learners are working individually on answering to some questions related to a theme they have collectively chosen at the begin of the activity (among a list of predefined ones, for example in the ecology domain, “waste sorting”, “water”, “biodiversity”, “renewable energies”). Their progression from one question to the other depends on the correctness of the answer provided. When they have all answered the last question, they are allowed to begin to discuss together and with the teacher about the theme and the answers they have provided.

The adaptation issues to deal with

The context in which we consider the adaptation problem is the following: an activity corresponding to the scenario given above is running. This means at least that:

  • the operationalization/instanciation data have been settled:
    • a group of participants has been registered to be involved in that activity. They are identified and known. Roles have been attributed to them so that they can play the scenario through these roles;
    • the resources and services required have been selected and are available;
  • the activity has begun;
  • there is someone (a teacher for example) who has the means of observing the activity and being aware of what is going on.

Adaptation issues to be considered by contributors may include but are not restricted to the predefined and "on the fly" adaptation. This is detailed in what follows.

Predefined adaptation

This is the case for foreseen events (adaptation is specified within the script)

Adaptation of the learning flow

The learning flow may not be the same for all the actors involved in the activity, even if they have the same role. It may also depend on what happens while the activity is going on.

Example in the case study: In the case study, there are at least two examples of predefined adaptation of the learning flow. The first one concerns the choice of the theme by the group of students at the beginning of the activity. The questions presented to them will depend on that theme. That means that, when preparing the activity, the scenario designer will have to define several sets of questions and to define to which theme each of these sets is related. Once the theme is chosen by the students, the associated set of questions will automatically be selected.

The answer-dependent progression from one question to another is the second example. It also has to be considered in advance and predefined by the scenario designer.

Adaptation “on the fly”

This is the case for unforeseen events (there is nothing in the script to define how to react). We have considered 3 situations in which adaptation « on the fly » may occur.

(1) Adaptation of the learning flow without modifying the scenario

Example in the case study: The students have chosen the theme to work on. They are currently involved in the quiz part of the scenario. Imagine that one of the students involved in the activity is late. He is still pondering on the first question whereas the other students are on the last one. The teacher then decides to let him forget about this first question and the intermediary ones and directly deal with the last one. The concerned student will thus “jump” from question 1 directly to last question.

(2) Modification of the operationalisation/instanciation data

This data is the one that is settled during the operationalization/instanciation phase, i.e. the data refers at least to the persons participating in the activity and the required resources (learning objects and services). They can be considered as the contextualization parameters. They are necessary to transform a scenario into an actual running activity. Modifying this data is at least a matter of:

  • modifying the participants currently involved in an ongoing activity (suppressing one of them - for example because s/he moved house and s/he now goes to a new school -, adding a new participant, or changing the distribution of roles, ...)
  • modifying the resources put at the participants' disposal

Example in the case study: The activity has begun. The students have chosen a theme and are currently answering to the questions. Imagine that a new student, Gerry, wants to join this group. What about adding him to this ongoing activity whereas he was not supposed at all to participate? Two cases could be distinguished: the one where the students are currently answering to the questions (they are in the quiz part of the scenario) and the one where they have finished the quiz part and are currently involved in the discussion part.

Note that this situation is different from the previous one (Situation (1) described above). In this case, the student to be added as a participant to the activity was not supposed to participate. In the previous one, he is a member of the group; he is involved in the activity; he is just late in doing the work.

(3) Modification of the scenario

Some events occurring within an ongoing activity may lead to the modification of the scenario itself.

Example in the case study: The students are currently involved in the discussion part of the activity. The teacher has the opportunity of inviting an expert of the theme to the discussion. He thinks this will be of a great interest for the students. To do that, he has to modify the scenario in order to add a new role (the “expert” one) and the means for this expert to join the discussion and participate to it. Also, he has to allocate this role to the expert he knows.

The related observation problems

A proposal for different purposes of observation:

  • Observation to be used by the engine for pre-defined adaptation: e.g. the selection of topic and answers to questions
  • Observation to be used by a moderator during the learning experience (on-the-fly adaptation): e.g. obeservation of the progress state of students, composition of groups (maybe a dropout or latecomer)
  • Observation to be used by analyst (a posteriori for research / evaluation): e.g. were there situations of “Conflict/disagreement” in the discussion?