Social XAI

Explainable Artificial Intelligence (XAI) describes methods that aim to make the behavior and decisions of AI more explainable, such as Shapley values. Social XAI puts focus on the social interaction between explainers and explainees within the process of explaining. The NLP Group studies how to generate explainee-specific natural language explanations and how to lead co-constructive explanation dialogues.

• Wachsmuth et al. (2026). Henning Wachsmuth, Kirsten Thommes, and Milad Alshomary. Operationalizing Social Interaction. In: Social Explainable AI, Springer 2026.
• Sengupta et al. (2025). Meghdut Sengupta, Maximilian Muschalik, Fabian Fumagalli, Barbara Hammer, Eyke Hüllermeier, Debanjan Ghosh, and Henning Wachsmuth. 2025. Investigating the Impact of Conceptual Metaphors on LLM-based NLI through Shapley Interactions. EMNLP Findings 2025.
• Fichtel et al. (2025). Leandra Fichtel, Maximilian Spliethöver, Eyke Hüllermeier, Patricia Jimenez, Nils Klowait, Stefan Kopp, Axel-Cyrille Ngonga Ngomo, Amelie Robrecht, Ingrid Scharlau, Lutz Terfloth, Anna-Lisa Vollmer, and Henning Wachsmuth. 2025. Investigating Co-Constructive Behavior of Large Language Models in Explanation Dialogues. SIGDIAL 2025.
• Alshomary et al. (2024). Milad Alshomary, Felix Lange, Meisam Booshehri, Meghdut Sengupta, Philipp Cimiano, and Henning Wachsmuth. 2024. Modeling the Quality of Dialogical Explanations. COLING 2024.
• Wachsmuth et al. (2022). Henning Wachsmuth and Milad Alshomary. 2022. “Mama Always Had a Way of Explaining Things So I Could Understand”: A Dialogue Corpus for Learning to Construct Explanations. COLING 2022.

• Alshomary (2023): Milad Alshomary. Audience-Aware Argument Generation. December 20, 2023. **Dissertation Award of Paderborn University**

• DFG-TRR 318 Subproject C07 (since 2026): Co-construction-following Large Language Models for Explaining
• NDS Hybrint  Subproject SP7 (since 2023):  Human-Centered Explanations of Machine Learning Results
• DFG-TRR 318 Subproject C04 (since 2021): Metaphors as an Explanation Tool
• DFG-TRR 318 Subproject INF (2021–2025): Towards a Framework for Assessing Explanation Quality