Four pillars of artificial intelligence research for education

Jill-Jênn Vie, a research fellow in the Soda project team at the Inria center in Saclay, wonders how artificial intelligence will transform the learning experience.

Far from having a definitive answer, he has identified four areas for analysis and research:

1. Fairness, confidentiality, and transparency of decisions.
   
   It's important to ensure that AI algorithms don't amplify existing biases, and instead go in the direction of reducing inequalities.
   

   "Many players go along with the CNIL's desire to minimize data, and wish to exclude the gender variable in their AI systems, but on the contrary we need to be able to measure discrimination in order to be able to reduce inequalities."
   

2. Useful metrics for teachers and students (learning analytics).
   
   To be able to give feedback to the learner so they know where they are in the knowledge space and how they are progressing.
   
   
3. Predicting student performance.
   
   To be able to intervene upstream, detect students in difficulty and adapt teaching accordingly.
   
   
4. Automatic content generation: written production, exercises or corrections.
   
   We have a real opportunity to generate exercises adapted to students' difficulties.

We often think of the risks, without considering the many opportunities that AI can bring to education and training.

For the full article: The four pillars of AI research for education

Illustration : Rawpixel - DepositPhotos

Visit inria.fr