🇫🇷Cabrilog

Use case 1 : Smart Enseigno AI

Smart Enseigno has an extensive collection of pedagogical resources covering 80 % of the cycle 2 mathematics program in France.

To use this collection, teachers are assisted by an AI service:

• in the fine analysis of students’ knowledge ;

• in the construction of learning paths.

The detailed analysis of the students' knowledge helps the teacher to identify weak knowledge or blocking difficulties for the student.

The creation of learning paths is used for mediation, to consolidate certain knowledge and to achieve new learning objectives set by the teacher. This service is aimed directly at the teacher and gives them the freedom to personalize and adapt the suggestions of the IA service. The use case as described above has been deployed and used for more than 3 years by nearly 15 000 users in France. This use case is therefore operational and able to produce usable data.

For this use case, it would be relevant to observe different data sets in order to progress in the creation of new learning paths.

In the current state, the AI service could be improved in the creation of paths to achieve a given learning objective. On a technical level as much as on a didactical level, it would be relevant to observe the choices made by teachers via various data sets.

We are looking for: learning traces data sets, learning analytics and adaptive learning solutions.

Use case 2 AI to support the student

This use case is positioned in a perspective of autotome use by the student :

• in a school context where the student will have access to a remedial service at home or in the school in total autonomy ;

• in a context of lifelong training allowing access to new skills, with a view to returning to work, training or professional reorientation.

The challenge is to be able to produce, from level diagnoses and learning objectives, learning paths covering various knowledge graphs.

In this use case, the challenge is therefore the creation of a system capable of analyzing traces of learning such as typologies of errors, independently of the chosen knowledge graph. In this use case, the challenge is both didactic and technical. Research in didactic faces a conceptual obstacle to describe independently of the task carried out, an knowledge graph.

The technical aspect of the choice of feedback by the AI service is linked to the didactic work. The didactic work brings the knowledge graph which is treated by the AI system to produce the feedback.

We are looking for: learning traces data sets, learning analytics and adaptive learning solutions.

Use case 3 AI to support scientist

This use case is aimed at the researcher. Learning traces are a powerful tool in many research fields : didactics of mathematics, cognitive sciences, developmental psychology, neurosciences, etc.

Learning traces are a powerful analysis tool. They allow the researcher to describe the interactions between a learner and mathematical concepts.

In this case of use, the researcher is able to produce digital supports for his experiments and to choose the learning traces to be collected (mathematical validity of a result, manipulation carried out, strategies for exploring a figure, eye-tracking, etc.).

In this case of use, the contribution of other AI systems is relevant and could allow the researcher access, for example, to new axes of analysis.

We are looking for: learning analytics and adaptive learning solutions.