Programme Empowering Learning Analytics

In his presentation, Hendrik Drachsler will introduce his current research program on Highly-Informative Learning Analytics (HILA). This program critically examines how learning analytics, artificial intelligence, and—in particular—generative AI (GenAI) can address the information and support needs of students and instructors in a meaningful, transparent, and evidence-based manner. The focus is on how AI-supported systems can not only generate automated information but also actually provide feedback that is effective for learning, understandable explanations, and adaptive support.

Hendrik will explain his perspective on Highly-Informative Learning Analytics, which was developed in collaboration with interdisciplinary experts from the fields of psychometrics, feedback theory, learning design, and artificial intelligence in education. Through experimental intervention studies in higher education and school settings, the research program investigates how AI-based feedback and support systems must be designed to ensure they are pedagogically meaningful, trustworthy, and sustainable.

As part of this research program, the so-called HILA-Manufaktur was developed—an evidence-based design methodology for creating Data-Enriched Learning Activities (DeLAs). DeLAs are data-enriched learning activities that can be directly applied in educational practice; they provide structured support for learning processes and, , simultaneously generate high-quality data for the creation of HILA feedback. Newer generations of DeLAs are increasingly integrating generative AI to generate individualized formative feedback, analyze learning products, stimulate reflection processes, and explain complex learning analytics dashboards in a way that is understandable to students and instructors.

On the one hand, DeLAs are characterized by their high flexibility, as they can be used in various disciplines and learning scenarios. On the other hand, they create stable and ecologically valid research conditions for systematically investigating factors that promote or hinder the use of AI-supported feedback and analytics systems. A particular focus is on the question of how GenAI can be meaningfully combined with learning analytics without compromising pedagogical quality, transparency, or data protection.

The goal of the HILA research program is to build an evidence-based knowledge base regarding the conditions for effective AI applications in educational processes. The presentation concludes with a critical analysis of the societal, technological, and pedagogical challenges associated with the implementation of generative AI, learning analytics, and intelligent assistance systems in higher education. In particular, it addresses issues of AI literacy, trust in AI systems, the explainability of analyses, as well as ethical and regulatory frameworks.

This workshop explores that question through a mix of short presentations and hands-on activities. We will look at how data can provide new insights into how students learn, where they struggle, and how they engage with course materials. We will also explore how learning analytics connects to learning design: how course structure shapes what data can tell us, and how data can feed back into better design decisions. Participants will try out tools and reflect on what meaningful instructor-facing applications look like in different institutional contexts.

What does a data infrastructure look like that truly accommodates the heterogeneous data sets of a university—ranging from campus management and LMS data to research and administrative data? This workshop addresses this question through a mix of short keynote presentations and hands-on exercises. We’ll start by clarifying the key building blocks of modern data architectures: What distinguishes a data warehouse from a data lake and a lakehouse, and when does each approach make sense? We’ll examine ETL and ELT pipelines, data modeling (such as star and snowflake schemas), data integration, and the roles of staging, core, and mart layers. In addition, we’ll address the practical framework conditions at the university: data quality and governance, metadata and data lineage, interfaces between source systems, as well as data protection and security requirements that shape every architectural decision. Participants will test tools and modeling approaches and reflect on what a sustainable, interoperable data architecture might look like in various institutional contexts.

Student-facing dashboards are everywhere, showing learners their progress, activity, and predicted outcomes. Yet the same dashboard is often read in very different ways by the students looking at it, and what gets shown is frequently what the data makes easy rather than what learners actually need. Designing student-facing learning analytics is non-trivial. In this hands-on workshop, we look at the why, what and how of effective learner-facing dashboard design. Participants will leave with a set of principles for grounding dashboards in feedback theory and learner needs.

The rapid expansion of artificial intelligence, learning analytics, and digital platforms is fundamentally reshaping education across schools, universities, and lifelong learning systems (Rienties et al., 2026). Yet much of the current discourse surrounding digital transformation remains dominated by narratives of efficiency, automation, optimisation, and scale. In this talk, Professor Bart Rienties argues that education risks losing sight of its fundamentally human purpose if digital futures are designed primarily around what can be measured, predicted, or automated.

Drawing on recent research on AI digital assistants, learning analytics, and large-scale online learning environments, including the development of institutional AI assistants designed to support learners ethically and responsibly at the Open University, this keynote explores how universities can move beyond “standard scripts” of technological adoption. Rather than positioning AI merely as a productivity tool, the talk examines how digital technologies might instead foster empathy, critical thinking, creativity, dialogue, and more inclusive forms of lifelong learning (The Open University, 2026).

Using examples from recent interdisciplinary research and large-scale studies in digital education, the talk reflects on emerging tensions between innovation and institutional responsibility, personalisation and surveillance, and efficiency and human agency. Particular attention will be paid to how learners, educators, and universities can collaboratively shape digital futures that are not only technologically advanced, but also socially just, emotionally intelligent, and environmentally responsible.

Rienties, B., Coughlan, T., Domingue, J., & Herodotou, C. (2026). New systems of learning for distance learning institutions? A six-study review of implementing AIDA. Computers and Education: Artificial Intelligence, 10, 100607. https://doi.org/https://doi.org/10.1016/j.caeai.2026.100607

The Open University. (2026). 5 factors influencing the future of global digital education.

AI and Digital-Supported Learning Processes as a Strategic Issue? Findings from the HFD Monitor 2026

Digital technologies and AI are transforming how learning is organized, supported, and assessed. This brings with it the hope of making learning processes more personalized, flexible, and efficient. As part of the HFD Monitor, vice presidents and provosts responsible for teaching were surveyed in May and June 2026 on the status of strategic engagement with digital and AI-supported learning processes: Where do higher education institutions stand today in the development and implementation of digital learning scenarios? What role do personalization, learning analytics, and AI assistants play in strategic processes?

Is Big Tech the only option? Learning Analytics as a Lever for Digital Sovereignty

Learning analytics reveals structural dependencies: Whoever collects and analyzes learning data determines the providers, infrastructures, and algorithms. Digital sovereignty and a transparent data culture are therefore closely intertwined. Without traceable systems, there can be no informed self-determination. How is the principle of digital sovereignty put into practice at German universities? The presentation focuses on governance, competencies, and open-source alternatives that make learning analytics controllable. The goal is to shift from reactive compliance to actively shaped sovereignty.