- The paper presents an interdisciplinary agenda linking AI, HCI, and learning sciences to create adaptive and immersive educational interfaces.
- It details a workshop model combining data-driven student modeling, user-centered design, and rigorous pedagogical evaluation for impactful learning tools.
- The proposed framework emphasizes collaborative research, ethical data practices, and scalable personalization to enhance learner agency and feedback.
Designing and Evaluating Next-Generation Learning Interfaces: Interdisciplinary Synthesis Across AI, HCI, and the Learning Sciences
Introduction
The increasing sophistication of generative AI, immersive environments, and multimodal interfaces has catalyzed a transformation of the learning landscape across educational, professional, and informal contexts. This workshop proposal, "Designing and Evaluating Next-Generation Learning Interfaces: Linking AI, HCI, and the Learning Sciences" (2604.25721), articulates a research and community agenda to synthesize advancements from artificial intelligence, human-computer interaction, and the learning sciences, driving forward the design of educational technologies that are at once technically robust, human-centered, and pedagogically grounded.
Motivation and Scope
Emergent AI systems are reshaping instructional content creation, adaptive tutoring, and collaborative knowledge construction. Systems leveraging LLMs, multimodal sensing, and augmented/virtual reality now support on-demand tutoring, learner-agency-driven exploration, and new modalities of collaboration. However, the effective deployment of such technologies in learning environments necessitates an interdisciplinary approach. Singularly focused advances—in adaptive modeling, interface design, or pedagogical theory—have historically failed to yield comprehensive educational impact without integration.
Prominent intelligent tutoring systems (ITS) demonstrate measurable learning gains by coupling data-driven student modeling, cognitive theory, and iterative HCI design. Yet, these systems are typically constrained by fixed interaction paradigms, which limit their capacity to address open-ended, multimodal, and collaborative learning scenarios enabled by contemporary AI and immersive technologies. The workshop confronts this gap by orchestrating a dialogue at the intersection of AI, HCI, and learning sciences, with a focus on addressing real-world learning challenges through interdisciplinary methods.
Thematic Foci and Research Questions
Three primary workshop aims structure the inquiry:
- Engaging AI and HCI researchers in educational design challenges: The workshop promotes the deployment of intelligent interfaces, immersive systems, and generative models in authentic learning contexts, emphasizing practical challenges in classrooms and workplaces.
- Surfacing learning science theories and evaluation methods: By integrating frameworks such as student modeling, educational data mining, and learning analytics, the event foregrounds theoretically rigorous and evidence-based design approaches, moving beyond purely technical or usability-driven perspectives.
- Bridging community silos for interdisciplinary innovation: The agenda is designed to stimulate new collaborations across disciplinary boundaries, identify shared research problems, and surface actionable principles for design and evaluation of next-generation learning interfaces.
Key discussion themes center on the controllability of AI in educational interfaces (addressing human-in-the-loop requirements), scalable feedback and personalization in large cohorts, data governance and privacy in learning analytics, and synthesizing divergent evaluation paradigms from HCI and education.
Program Architecture and Expected Contributions
The planned workshop is highly interactive, centering on:
- A keynote framing current challenges at the intersection of AI, HCI, and learning sciences.
- Spotlight presentation sessions for selected submissions, with structured cross-disciplinary feedback to identify methodological synergies and research gaps.
- Thematic breakout discussions on control affordances, feedback mechanisms, data/ethics, and evaluation strategies.
- Structured speed collaboration rounds to facilitate new partnerships and ideate on integrative research directions.
Through this format, the workshop is expected to yield an actionable interdisciplinary research agenda; early-stage pilot collaborations; and an overview of design, evaluation, and deployment principles for learning interfaces that transcend the limitations of traditional ITS and HCI silos.
Context With Prior Work
The community lacks sustained venues that fuse AI modeling, user-centered interface design, and rigorous pedagogical research. While flagship conferences such as NeurIPS, AAAI, KDD, CHI, ICLS, and LAK have hosted workshops on AI for education, augmented educators, and LLM applications in qualitative research, these efforts remain fragmented along disciplinary boundaries. This workshop differentiates itself by embedding interdisciplinary synthesis into its structure and outputs.
Foundational work in cognitive tutors, data-driven student modeling, and the integration of HCI methods in EdTech (2604.25721, Zhang et al., 6 Feb 2026) underscores the utility and limitations of purely siloed approaches. Addressing issues such as agency in AI-guided learning, scalable personalized feedback, and reconciling HCI-centric and educational evaluation metrics demands collaborative approaches that are a central focus for the proposed workshop.
Theoretical and Practical Implications
The integration of AI, HCI, and the learning sciences has the potential to significantly enhance the controllability, explainability, and adaptation of educational interfaces. This multidisciplinary convergence can yield systems that:
- Dynamically calibrate automation to preserve learner agency and educator oversight;
- Deliver scalable, context-rich, and actionable feedback at both individual and cohort levels;
- Balance fine-grained learning analytics with ethical standards around privacy and transparency;
- Adopt evaluation methodologies that address learning outcomes, usability, and equity of access.
Strong claims are made about the necessity of explicit interdisciplinary intersections to move beyond the constraints of monolithic AI or HCI-driven design paradigms. The workshop anticipates that such an approach will enable more robust, scalable, and contextually adaptive learning systems.
Future Directions
Future research pathways identified include:
- Development of composable, transparent, and ethically stackable educational interface architectures that operationalize findings from HCI, AI, and education.
- Longitudinal studies on agency, efficacy, and engagement in AI-mediated, immersive, and multimodal learning environments.
- Community toolkits and datasets supporting reproducibility and cross-domain benchmarking.
- The formulation of unified evaluation frameworks integrating educational impact metrics, UX principles, and AI alignment diagnostics.
Conclusion
"Designing and Evaluating Next-Generation Learning Interfaces: Linking AI, HCI, and the Learning Sciences" (2604.25721) provides a timely and rigorous agenda for unifying advances from AI, HCI, and pedagogy to inform the design, evaluation, and deployment of educational technologies. By convening interdisciplinary discourse and catalyzing collaboration, the workshop advances both theoretical understanding and practical strategies for a new class of learning interfaces characterized by adaptability, human-centeredness, and pedagogical alignment.