Research Associate / Doctoral Researcher in Natural Language Processing and Causal AI for Scientific Reasoning German National Library of Science and Technology (TIB) in Germany

German National Library of Science and Technology (TIB) in Hannover is offering a Research Associate / Doctoral Researcher position in Natural Language Processing and Causal AI for Scientific Reasoning within the Open Research Knowledge Graph (ORKG) and the Data Science and Digital Libraries Research Group led by Prof. Dr. Sören Auer.

The project is part of the Leibniz Collaborative Excellence project EcoAIRE, which combines computational NLP expertise at TIB, ecology at the Leibniz Institute of Freshwater Ecology and Inland Fisheries (IGB), and causal modeling expertise at the German Research Center for Artificial Intelligence (DFKI). The research aims to support AI-assisted hypothesis assessment in invasion biology and the restoration of freshwater ecosystems using heterogeneous scientific evidence.

The doctoral topic focuses on NLP for scientific and causal information extraction. The successful candidate will develop methods to transform scientific publications and other research artifacts into structured, machine-actionable representations of hypotheses, variables, context, uncertainty, and causal relationships. The work extends beyond fact extraction to modeling how evidence fits together and under which conditions claims hold. A key part of the position is exploring causal and world-model-inspired approaches to scientific reasoning, including simulation, intervention-based reasoning, and hypothesis assessment under changing conditions.

Tasks include building a multimodal, provenance-aware framework that integrates publications, images, supplementary information, datasets, and related artifacts into structured representations for hypothesis assessment; developing language-model and agentic methods for extracting, normalizing, aligning, and synthesizing scientific claims and evidence; investigating causal reasoning methods for interventions and generalization; and creating evaluation frameworks using expert-grounded assessment and LLM-as-a-judge paradigms. The role also involves publishing in peer-reviewed NLP, Semantic Web, AI, and machine learning venues and contributing to project reporting and dissemination.

The profile requires a completed Master’s degree or equivalent in computer science or a closely related field such as AI, NLP, machine learning, data science, or computational linguistics. Applicants should have practical experience with Python and modern software engineering practices, hands-on work with transformer-based models and large language models, and experience with agentic or multi-agent AI workflows. A very good command of English at C1 level is required. Desirable experience includes LLM fine-tuning, advanced RAG pipelines, causal graphs or structural causal models, Semantic Web technologies, and domain knowledge in ecology, especially freshwater ecology, invasion biology, or restoration ecology.

The position is fixed-term for three years, full-time at 39.8 hours per week, and generally suitable for part-time work. Remuneration is based on TV-L E13. Mobile work is possible to a certain extent. TIB notes that the appointment is subject to final funding approval by the Leibniz Association. The post includes a scientifically and intellectually inspiring environment, a modern workplace in Hannover, flexible working options, training opportunities, supplementary pension coverage, and support for equipment, conference travel, and research visits.

Applications are open until 2026-11-14. Candidates should apply online via the application link or send a complete paper application / PDF with subject 44/2026. For foreign degrees, a Statement of Comparability from ZAB is required unless the degree is already recognized in ANABIN. Questions can be directed to Dr. Jennifer D’Souza by email.