Paderborn Colloquium on Data Science and Artificial Intelligence in School

LLMs cannot only generate text that is barely distinguishable from human-written anymore, but the most recent LLMs can even tackle problems successfully that they have never seen before. In this talk, I start from the general functionalities of LLMs, before I presented insights from selected research of my group involving LLMs to reconstruct, optimize, and create natural language text for specific NLP tasks. On this basis, I look at the recent breakthroughs that caused the success of ChatGPT-like technologies and the paradigm shift that comes with it.

After receiving his PhD from Paderborn University in 2015, he worked as a PostDoc at Bauhaus-Universität Weimar, before he returned to Paderborn as a junior professor from 2018 to 2022. His group studies how intentions and views of people are reflected in language and how machines can understand and imitate this with large language models. Henning’s main research interests include computational argumentation, the mitigation of social bias and media bias, and the construction of human-like explanations for educational and explainable NLP.

Some highlight the amazing potential of data’s power to create a better world and others see portents of how such use of data will only exacerbate inequities and lead to the very downfall of democracy. Regardless of what stance you take, being data literate is crucial, but it is unclear what kind of data literacy is important and for whom. In this talk I will lay the groundwork for taking a critical literacy perspective on data literacy to focus on reading and writing the word and the world with data. I will also spend time connecting such work to practice by discussing an ongoing research to practice partnership with teachers and share a framework for thinking about practices of reading the word and the world with data visualizations.

While working on a master’s degree, I realized not only how little I understood statistics, but that I had been doing a terrible job of teaching those concepts to students for years. It was this realization that motivated me to pursue a PhD in Mathematics Education to investigate the question of how can we improve the statistical education of students in schools? My work stems from two main principles: that education is transformative, and the goal of public education is democratic equality. More specifically, my research is focused on issues of equity, including the consideration of issues of access, achievement, identity and power, at their intersection with statistics and data science education.

The COVID-19 pandemic highlighted how mathematics and statistics can be used to present sometimes contradictory and misleading claims by various groups – amplifying the need for citizens be capable of critically evaluating claims made by both expert and non-expert commentators, and the decisions of government. This means that informed and responsible citizens must be capable of understanding the role of mathematics and statistics in underpinning evidence and have the capacity to employ evidentiary practices in forming relevant judgements. In this presentation I report on the aims and approaches that frame an international collaborative project being conducted by Australian Catholic University and Wurzburg University, entitled Strengthening Teachers’ Instructional Capabilities with Big Data. The project is designing and implementing tasks that require middle years students to evaluate differing claims in the media about issues related to sustainability by making use of relevant publicly available large data sets. This includes students’ selection of databases, approaches to modelling data, and their decision-making processes. Participant teachers will engage with professional learning activities based on students’ approaches to resolving differences in media reports.

He is the Research Director for the STEM Education Program – an interdisciplinary research space focused on the enabling and transformative role of mathematics within the STEM disciplines. His research interests span across the areas of critical mathematical thinking, numeracy, mathematical modelling, the use of digital resources in mathematics teaching and learning, and the mathematical/statistical capabilities required for informed and responsible citizenship. This work is driven by awareness that the capacity to know and use mathematics confidently is important for an individual’s career prospects and their empowerment as informed citizens.

In this presentation, I will describe our creation of CMSE 201, „Introduction to Computational Modeling and Data Analysis,“ which is intended to be a standalone course teaching undergraduate students (both STEM and non-STEM) core concepts in data analysis, data visualization, and computational modeling. I will discuss the rationale behind the „flipped classroom“ instructional model that we have been using and explain the course’s design principles and implementation. The concepts and skills students learn in this course can be used by other disciplines as the foundation for integrating computing across the curricula in undergraduate degree programs.

He received his PhD in Astrophysical and Planetary Sciences from the University of Colorado where he worked as a computational astrophysicist running simulations aimed at understanding various facets of chemical evolution in the universe. He then worked as a National Science Foundation Astronomy and Astrophysics postdoctoral fellow at MSU before joining the ranks of the CMSE faculty in 2017. Most recently, Devin has been pursuing computing education research in an effort to better understand how students learn to do computational and data science and help inform pedagogical approaches in these disciplines. As part of this effort, Devin co-leads the Computing Education Research Lab at MSU.