Towards Sensor-based Learning Analytics: Exploring the Use of Multimodal Data to Understand Students’ Learning Processes

Dra. Namrata Srivastava

Vanderbilt University, EE.UU.

Resumen/Abstract

When students learn, much of their thinking and interaction happens beneath the surface, visible neither in their final answers nor in classroom observations. My research explores how sensor technologies can uncover these hidden dynamics and help us understand learning as it unfolds. I began this work during my Ph.D., using thermal cameras and eye-trackers to explore how patterns in gaze and thermal signals reveal students’ attention and cognitive load during video-based learning. These early studies demonstrated how multimodal data can capture subtle indicators of cognitive and affective states. Building on this foundation, my current work extends these ideas to K-12 classrooms, where learning is social and dynamic. By combining data from sensors, classroom interactions, and student activity logs, I examine how students regulate their learning and collaborate during inquiry-based science and engineering tasks. This research contributes to developing classroom-ready multimodal learning analytics pipelines that can help teachers interpret learning as it happens and create more responsive and inclusive environments for all students.

Profesora Proponente EPS: Ruth Cobos.

Curriculum ponente

Dr. Namrata Srivastava is a Research and Development Scientist at the Institute for Software Integrated Systems (ISIS), Vanderbilt University, where she works on human-centered AI and multimodal learning analytics to enhance equitable and responsive teaching and learning in STEM+C classrooms. Her research focuses on understanding students’ inquiry processes, collaboration, and learning performance through multimodal data such as eye-tracking, dialogues, and interaction logs. Prior to this, she was a Postdoctoral Researcher at the University of Pennsylvania’s Penn Center for Learning Analytics, developing AI-based detectors of student engagement (e.g., boredom, confusion). She also serves as Adjunct Research Fellow at Monash University, Australia  investigating self-regulated learning and engagement in digital learning environments.Dr. Srivastava earned her Ph.D. in Computer Science from the University of Melbourne, where she pioneered research on sensor-based learning analytics to detect cognitive load using non-invasive physiological sensors. Her work has contributed to major international projects such as NSF EngageAI, SPICE, and FLoRA, and collaborations with institutions like Stanford University, Adobe Research, and TU Delft. A recipient of the 2024 Emerging Scholar Award from the Society for Learning Analytics Research (SoLAR), her interdisciplinary research bridges AI, HCI, and Data Science to design intelligent systems that foster more effective, inclusive, and data-informed education.