Selected finalist projects to be presented
June 22nd – h: 12:10 – 13:10 CET
• Denise Flämig and Jule Kuppinger
(University of Education Weingarten)
“Exploring Climate Decision-Making through VR: Navigating Urban Heat”
Abstract: The project presents an interactive virtual environment in which users navigate through a simulated urban setting designed with a focus on climate-friendly principles. Participants move through the city by following guiding arrows that lead them sequentially from one designated measurement point to another. At each location, users are tasked with assessing localized temperature conditions, thereby engaging with the spatial variability of urban heat. By choosing between alternative climate adaptation measures and seeing their simulated impacts in real time, participants experience the complex trade-offs between cooling, energy, costs, and urban quality of life that form the core of the environment.
• Mariana Madalina Nastase
(University of Craiova)
ChoiceCrafter: Integrating Behavioural Nudging into a Mobile Microlearning Ecosystem for Smart Education
Abstract: This work introduces ChoiceCrafter, a mobile learning ecosystem that integrates microlearning with behavioural nudging techniques. The system is designed to support both students and instructors through a dual-interface architecture that combines structured learning experiences with subtle behavioural interventions. The central idea behind ChoiceCrafter is that engagement should not be left to chance. Instead, it can be intentionally supported through carefully designed mechanisms that encourage continuity, attention, and self-regulation
• Alexandru Smarandache
(University of Craiova)
Quizora: An AI-powered platform for automated assessment in education
Abstract: Quizora is an educational platform designed to automate the formative
assessment lifecycle within Smart Learning Ecosystems. The core problem addressed is the ‘feedback bottleneck,’ where time constraints prevent instructors from providing personalized evaluations to large numbers of students. The proposed solution integrates Google’s Gemini 3.1 Pro to evaluate open-ended student responses against instructor-defined rubrics, ensuring pedagogical rigor through a human-in-the-loop workflow. To sustain motivation, the system employs gamification mechanics grounded in Self-Determination Theory, such as competence badges and habit-forming streaks
• Yara Youssef
(DIPF | Leibniz-Institut für Bildungsforschung und Bildungsinformation Leibniz Institute for Research and Information in Education)
LearnBuddy: Supporting Self-Directed Learning through LLM-Generated Quizzes from Personal Study Materials
Abstract: LearnBuddy is a web-based study application that allows learners to upload their course notes and automatically generate personalized practice content from that material. It offers three core interaction modes.
1) Quiz mode: the student selects one or more document sets and a question count. The system generates a mix of multiple-choice and free-text questions grounded in the uploaded content, then provides explanatory feedback on each answer.
2) Exam mode: once a student scores 75% or above in practice, they can attempt an exam version; same format, but without mid-session feedback, simulating real assessment conditions.
3) Content Lab: a free-form chat interface where the assistant is fully aware of all uploaded documents, enabling targeted explanations and additional exercises drawn from the student’s own material.