/AIDA

AIDA

AI-Driven e-Assessment

Smart education logo

Creating exercises and tests is a heavy burden for teachers that could be lifted using artificial intelligence. However, challenges remain. Creating personalized questions at the right difficulty levels is technically complex, for instance, and user adoption – for teachers and students alike – isn’t always a walk in the park. The AIDA project will tackle these and other challenges by conducting further research into e-learning, e-assessment, system usage visualization, natural language processing and transparent AI models. The project will also create an AI-driven tool for evaluation in secondary school e-assessment and e-learning in higher education.

AI-driven teaching assistants?

Formulating test questions in educational settings requires training, experience, time and resources. Even more, in high-stakes settings like certifications and tests, questions cannot be reused. However, many digital questions and teaching materials are already available, sparking the idea of using AI techniques to support the creation of questions and exercises and offer personalized, actionable insights to students and instructors.

Uncertain results, a steep learning curve

Creating an AI-driven support system for educators is challenging for a number of reasons. For example, little research has been done on the automated generation of questions based on existing course material – particularly in minority languages. In addition, the user adoption of such a system is complicated, as few teachers are familiar with the design, configuration and limitations of AI systems.

A technically validated, easy-to-use, interactive system

Based on a strong partnership of industrial and academic partners, the AIDA consortium will explore applications in e-assessment in education, including an AI support component. Partners will build a proof of concept that will be tested in real-life secondary schools and aims to achieve four innovation goals:

  • efficient AI-supported question generation;
  • AI-driven e-learning personalization through an interactive interface;
  • increased user adoption;
  • insights mining for non-technical users, such as teachers.

More effective learning – and teaching

The outcomes of the AIDA project will expand consortium members’ offerings in the domains of e-learning, user engagement, automated insights, distance learning, student engagement and personalized learning solutions. The proof of concept will pave the way for cutting-edge educational approaches tailored to the individual needs of students, while offering teachers efficiency-boosting, insightful and effective support tools.

“The AIDA project will tackle these and other challenges by conducting further research into e-learning, e-assessment, system usage visualization, natural language processing and transparent AI models. The project will also create an AI-driven tool for evaluation in secondary school e-assessment and e-learning in higher education.”

AIDA

AIDA aims to design and evaluate new explainable AI methods to create material for online learning and e-assessment.

 

AIDA is an imec.icon research project funded by imec, Innoviris and Agentschap Innoveren & Ondernemen.

 

It started on 01.10.2020 and is set to run until 30.09.2022.

Project information

Industry

  • Cumul.io
  • Televic Education
  • Nextbook
  • Wezooz

Research

  • imec – IDLab IBCN – UGent
  • KU Leuven – HCI

Contact

  • Project lead: Filip Vanlerberghe
  • Research lead: Thomas Demeester
  • Proposal Manager: Thomas Demeester
  • Innovation manager: Eric Van der Hulst