cubes COMPetences and PERsonnalization
Project ANR-18-CE38-0012
WP head : LIRIS (N. Guin and M. Lefevre)
Participants : teaching teams from the 3 fields, head of WP1 and WP3
Objectives :
Once the skills to be acquired have been represented in a reference framework, the objective is to be able to propose activities to the learner that mobilise these skills and to analyse the traces of these activities in order to evaluate how the learner masters each of these skills.
With regard to the creation of activities, we have the authoring tool ASKER, developed at LIRIS, which enables a teacher to define an exercise model that will then be used by an exercise generator to generate a large number of exercises to work on the same skill. This authoring tool enable to generate exercises such as open-ended and short answer questions, multiple choice questions, matches, categorizations, schedules, and can therefore be used in any field. However, the knowledge of the field and the skills mobilized by the exercises are not currently represented in ASKER. The first scientific lock of WP2 will therefore consist in associating to an exercise model the competence(s) mobilized by the exercises resulting from this model. Another research question will be to use the information on the competencies present in the competency framework to vary the competencies and the difficulty of the exercises generated.
Once the learner can be offered exercises on the skills to be acquired, another scientific lock is to analyse the learner's responses to these exercises in order to assess whether he or she masters the skill(s) being mobilized. This requires having traces of activities that respect a model that can be applied to various activities and integrate the skills mobilized by the activity. The model for calculating the mastery rate of each skill should take into account the evolution over time of the learner's responses to exercises involving this skill, but also any information related to the difficulty of the exercise in relation to this skill. This calculation model will also have to take into account the prerequisites relationships between skills represented in the meta-model of reference: when an exercise is successful, we can positively impact not only the skill mobilized by the exercise, but also the skills pre-required for that skill. Finally, this profile development mechanism should make it possible to take into account human opinions on the mastery of skills assessed outside the traced IT environment.