We are proud to announce the recent publication and conference presentation of a new work, “Graph-Based Real-Time Dialogue Adaptation for Personalized Virtual Job Interview Training”, led by Deeksha Adiani. Deeksha presented this work at the International Conference on Graph Transformation (ICGT) 2025 in Koblenz, Germany.
This research tackles a growing challenge in digital learning: how to create flexible, adaptive systems that respond meaningfully to the needs of diverse users. Many e-learning platforms rely on rigid, pre-scripted dialogues that fail to reflect real-world communication, particularly in high-stakes scenarios like job interviews. This work introduces a graph rewriting framework that enables dynamic adaptation of dialogue based on real-time multimodal input—including speech, eye gaze, and physiological signals such as heart rate.
By simulating interview scenarios in a more personalized and inclusive manner, this approach has strong implications for supporting neurodiverse learners, including autistic individuals who may face social communication barriers during interviews.
🔗 Full Paper: https://doi.org/10.1007/978-3-031-94706-3_11