Generative AI systems such as large language models are powerful but often opaque, producing answers without clear justification. XAI-CA addresses this challenge by developing methods that combine the logical rigor of computational argumentation with the expressive capabilities of neural language models. The goal is to create hybrid neurosymbolic approaches that generate explanations people can understand and trust.
Planned use cases include AI-assisted legal reasoning, research integrity in publishing, and industry decision support, demonstrating how argumentation-based explanations can improve the accountability of AI across domains. By uniting expertise in computational argumentation, symbolic AI, and data provenance, the project aims to advance explainable and trustworthy AI methods.
Principal Investigators: Bertram Ludäscher (Illinois), Martin Caminada (Cardiff)
Funding Period: 2024–2025
Funding Source: University of Illinois System / Cardiff University Joint Seed Grant (Cardiff–DPI Science Teams initiative)
Partner Institutions: Cardiff University
How can we trust the integrity of results from research that relies on computations without repeating them? By certifying the successful original execution of a computational workflow that produced findings in situ. With certifications in hand, consumers of research can trust the transparency of results without necessarily repeating computations. The TRACE project is developing techniques and tools for certifying the results of computational research. This includes the creation of technical specifications for describing and certifying research artifacts as well as the development of reusable software tools
Principal Investigators: Bertram Ludäscher, Timothy McPhillips, Kacper Kowalik, Craig Willis
Funding Period: July 2022 – June 2025
Funding Source: NSF (2209628)
Partner Institutions: Cornell University, University of North Carolina at Chapel Hill
Whole Tale is an NSF-funded Data Infrastructure Building Block (DIBBS) initiative to build a scalable, open source, web-based, multi-user platform for reproducible research enabling the creation, publication, and execution of tales – executable research objects that capture data, code, and the complete software environment used to produce research findings. A beta version of the system is available at https://dashboard.wholetale.org. The Whole Tale platform has been designed based on community input primarily through working groups and collaborations with researchers.
Principal Investigators: Bertram Ludäscher, Matthew Turk
Funding Period: March 2016 – February 2023
Funding Source: NSF (1541450)
Partner Institutions: University of Chicago, University of Texas at Austin, University of Notre Dame, UC Santa Barbara, University of Southern California, Cornell University, University of North Carolina at Chapel Hill