MAAI4Ed: Multimodal Affect in AI for Education

Design, Application, and Ethical Implications
AIED Workshop (Half-day)

Location TBA

→ Submit a Paper

Overview

As artificial intelligence becomes increasingly embedded in educational contexts, the ability of AI systems to perceive, interpret, and respond to learners' affect has shifted from a niche research interest to a central necessity. While traditional research has prioritized "cold" cognitive processes, affect represents the "hot" functional dimension that is intrinsically intertwined with learning outcomes.

The Multimodal Affect in AI for Education(MAAI4Ed) workshop focuses on (1) AI-driven approaches in detecting and interpreting affective states through multimodal data streams and (2) the design of affect-aware AI systems that support emotional regulation and learner well-being. By fostering interdisciplinary dialogue among researchers, designers, and practitioners, the workshop aims to advance ethically responsible affect-aware AI systems for education.

Keynote Speaker

Keynote Speaker

Dr. Roger Azevedo

Feeling the Learning: Designing Affect-Aware AI Systems Through Multimodal Analytics and the Future of Emotionally Intelligent Systems

Learners do not simply think through complex educational experiences; they feel them. Yet, the social and emotional dimensions of learning remain among the most underutilized signals in AI-driven educational systems. Drawing on two decades of research examining cognitive, affective, metacognitive, and motivational (CAMM) processes, this keynote spans work across intelligent tutoring systems, serious games, and immersive learning environments. It then synthesizes key empirical findings that demonstrate how multimodal data streams, such as facial expression recognition, eye tracking, physiological sensors, and interaction logs, can detect, model, and respond to learners' affective states in real time. Rather than treating affect as a nuisance variable, this presentation positions emotion as central to AIED design. Findings from MetaTutor and similar systems reveal how co-occurring emotional states, like confusion-to-frustration transitions, interact with metacognitive accuracy and learning outcomes. These insights inform UX decisions, adaptive scaffolding strategies, and interface transparency. Finally, the keynote addresses critical challenges and presents a forward-looking vision for emotionally intelligent, human-centered learning systems.

Biography

Dr. Azevedo is a Pegasus Professor in the School of Modeling, Simulation, and Training at the University of Central Florida. He is also an affiliated faculty member in the Departments of Computer Science and Internal Medicine at the University of Central Florida and the lead scientist for the Learning Sciences Faculty Cluster Initiative. He received his PhD in Educational Psychology from McGill University and completed his postdoctoral training in Cognitive Psychology at Carnegie Mellon University. His main research area includes examining the role of cognitive, metacognitive, affective, and motivational self-regulatory processes during learning, reasoning, and problem solving with intelligent learning technologies such as intelligent tutoring systems, hypermedia, multimedia, simulations, serious games, immersive virtual learning environments, human digital twins, and simulated learners). He has published over 300 peer-reviewed papers, chapters, and refereed conference proceedings in the areas of educational, learning, cognitive, and computational sciences. He is a fellow of the American Psychological Association, the American Educational Research Association, and the recipient of the prestigious Early Faculty Career Award from the National Science Foundation. He was recently inducted into the Academy of Science, Engineering and Medicine of Florida.

Agenda

The MAAI4Ed workshop is formatted as a half-day session (approx. 4 hours) aimed at bringing together a diverse cohort of learning scientists, AI specialists, and UX designers.

  • Opening Remarks

    Time: 14:15 - 14:30

    Introduction to the MAAI4Ed vision and goals.

  • Keynote Speaker

    Time: 14:30 - 14:50

    Speaker: Prof. Roger Azevedo

  • Lightning Talks

    Time: 14:50 - 15:50

    "Lightning Paper" presentations to spark ideas.

  • Coffee Break

    Time: 15:50 - 16:00

    Networking and informal discussion.

  • Interactive Design Seminar

    Time: 16:00 - 16:45

    Hands-on group activity: "Designing the Ideal Affect-Aware Tool."

  • Plenary Discussion

    Time: 16:45 - 17:45

    Synthesizing themes: Design, Application,and Ethical Implications.

  • Wrap-up & Next Steps

    Time: 18:15

    Closing remarks and information on publishment.

Key Information

Submission deadline: TBA

Notification of acceptance: TBA

Workshop date: TBA (Half-day)

Workshop location: TBA

Contact: xiaoshan.huang@mail.mcgill.ca, Andy.Nguyen@oulu.fi

Call for Participation

We invite researchers, designers, and practitioners from the AIED, EDM, and Learning@Scale communities to participate. Attendees are encouraged to submit 4-8 page position papers detailing recent empirical research or theoretical frameworks regarding multimodal affective analytics in education.

The workshop's mission is structured around two central pillars:

  • Pillar 1: Detection and Interpretation of Affect in Learning Settings. Leveraging advanced AI techniques to decode complex affective states in real-time.
  • Pillar 2: Affect-Aware Design. Moving beyond detection to the creation of AI interventions that actively scaffold emotional regulation and promote learner well-being.

Submission Format: 4-8 page research(e.g., experimental, design-based), review, or position papers.

Review Process: These submissions will undergo a peer-review process by the organizing committee to ensure relevance and academic rigor.

Dissemination: To ensure a lasting impact, all accepted position papers will be published in a dedicated digital repository, such as the CEUR Workshop Proceedings (CEUR-WS).

Organizers

Xiaoshan Huang

Xiaoshan Huang

McGill University, Canada

Andy Nguyen

Andy Nguyen

University of Oulu, Finland

Jie Gao

Jie Gao

McGill University & Mila, Canada

Haolun Wu

Haolun Wu

McGill University & Mila, Canada; Stanford University, USA

Yimeng Wang

Yimeng Wang

Yale University, USA

Tony Ahn

Tony Ahn

University of British Columbia, Canada

Tiantian Jin

Tiantian Jin

Columbia University, USA

Roger Azevedo

Roger Azevedo

University of Central Florida, USA

Susanne Lajoie

Susanne Lajoie

McGill University, Canada