Use of Profiling to optimise a collaborative session in a learning game
Résumé
Our research work deals with the development of new learning environments, and we are particularly interested in studying the different aspects linked to users’ collaboration in these environments. We believe that Games-based Learning can significantly enhance learning. That is why we have developed learning environments grounded on graphical representations of a course. These environments allow us to set up experiments with students in our university. The emergence of online multiplayer games led us to apply the metaphor of exploring a virtual 3D world, where each student embarks on a quest in order to collect knowledge related to a learning activity. In the environment, each part of the world represents a place, sometimes a collaborative place, where students are supposed to acquire a particular concept. Learning objects, artefacts or collaborative tools may be present in each location and a correct answer to specific exercises gives a key to the students, allowing them to access other activities.
Although the students appreciate this approach, there is an obvious need for information about students’ skills, especially for the teacher. Indeed, the teacher needs to adapt his/her pedagogical session according to what is going on during the collaborative activity. But, this adaptation only becomes possible when s/he has particular information concerning the users in the environment. In our approach, we can update the user model of the students thanks to specific questionnaires or behaviours in the world and use the results in order to enhance collaboration when such particular events occur, by using both data collected from traces resulting from the collaborative learning activity and information present in student’s user models.
In this paper, we focus on the need for collaborative information concerning the users in order to create and optimize teams. First, we describe briefly the “Learning Adventure” environment: a general game-based platform. Then, we explain how we update user model of the student via an adapted questionnaire (e.g. Belbin’s Team-Role Self Perception Inventory). Then, when particular events occur, we notify the teacher of possible lack or problem in team constitution in regard of particular aspects collaborative activity (e.g. a team has difficulties to propose innovative solutions, to respect delay, or some students are chatting too much). From this statement, it became crucial to have indicators concerning both the knowledge (success and failure for sub activities) and the behaviour (chatterer, cooperative, resource investigator, co-ordinator, implementer, shaper, specialist or expert) of the different students during the pedagogical session. These indicators help us to propose and to take the right decisions to enhance collaboration. We also propose innovative collaborative tools or artefacts directly in the game, making a complete immersion possible. A real experiment has been made at our university in the domain of project management validating the feasibility of the approach.