Junior fellows and distinguished dissertation of the GI & AI for crisis

In this issue, we introduce four recent Gesellschaft für Informatik (GI) junior fellows and highlight one distinguished dissertation. These scholars share insights into their research journeys and key contributions to their fields. Additionally, we present a response to our Call for Papers on AI for Crisis, demonstrating how AI-driven approaches can enhance resilience in times of uncertainty.

In his article “Natural Language Processing for Social Good: Contribution to Research and Society,” Daniel Braun illustrates how AI – especially NLP – is changing our world. His appointment as a GI Junior Fellow emphasizes the importance of navigating the responsibilities associated with the growing use of AI. Daniel is a professor in Marburg and highlights key aspects in his article, such as trustworthiness; in particular, trustworthy annotated data shapes the development of AI. His article effectively demonstrates how AI in law works and how we – as computer scientists – can influence society.

In the second article, Franziska Boenisch introduces herself in “A Self-Portrayal of GI Junior Fellow Franziska Boenisch: Trustworthy Machine Learning for Individuals” as part of our series on recent GI junior fellows. Franziska is a tenure-track faculty member at the CISPA Helmholtz Center for Information Security, where she co-leads the SprintML lab focused on secure, private, robust, interpretable, and trustworthy machine learning. The author highlights her perspective on user-centered machine learning privacy and how federated learning, differential privacy methods, and even single neurons leak private data. As AI increasingly integrates into our daily lives, safeguarding our privacy is more important than ever.

Mareike Lisker’s article aims to demonstrate how academic careers can be unpredictable and subject to change. She argues that increasing expectations for digital literacy burden users with managing their data, even though they are often ill-equipped due to pervasive tracking systems. Think about how little users know about cookies. The article then connects this topic to Lisker’s academic journey, tracing her path from her Master’s research to her current PhD project on content moderation in decentralized platforms. This shows us how our research field is not only diversifying its participants and topics but is also becoming increasingly interdisciplinary. This promises to make us, as researchers, more resilient in a rapidly changing world.

Bettina Finzel’s article, “Toward Trustworthy AI with Integrative Explainable AI Frameworks,” explores the pressing challenge of ensuring AI systems are reliable, transparent, and interpretable, particularly in high-stakes domains like healthcare. As AI regulations, such as the European AI Act, continue to shape the landscape, Bettina proposes an integrative Explainable AI (XAI) framework that merges interpretability, interactivity, and robustness to enhance human-AI collaboration. The article highlights methods for evaluating AI models, mitigating bias, and fostering interdisciplinary cooperation to create AI systems that are not only technically sound but also ethically and socially responsible.

The distinguished dissertation “Scalable SAT Solving and Its Application” by Dominik Schreiber, a Young Investigator Group Leader at KIT, focuses on simplifying and accelerating the process of solving SAT problems. The author notably leveraged powerful computing systems, such as supercomputers and cloud computing, to enhance these solutions. The author’s system, MALLOB, has established itself as a world-leading automated reasoning system. This article offers an overview of the research, its key findings, future impact, and some personal reflections from the author.

Finally, we add a paper on how to generate data for AI processes. As highlighted by some of the previous articles, one can assume that we live in times of crisis. Responding to our call for AI for crisis articles, Felix Brei et al. present their approach, “Queryfy: From Knowledge Graphs to Questions using Open Large Language Models. Large Language Models (LLMs)” which can assist with the translation of natural language into SPARQL queries. Thus, questions such as “Which of our suppliers are based in countries with ongoing armed conflicts?” can be answered, and hence, systems based on this dataset can drive the resilience of companies and organizations forward.

As the editor, I want to share a few words: My time with Information Technology has recently ended, and I am grateful that this positive special issue, which highlights how technology – particularly AI – can benefit the world, is my final editorial contribution. Thank you for reading Information Technology.

Sapere Aude!