Reproduce “Argumentative XAI: A Survey” on OpenResearch

Turn your favorite coding agents into research agents that review literature, develop hypotheses, and run experiments.

Universal app for Apple silicon and Intel. Requires macOS 11 or later.

About this paper

Kristijonas Čyras, Antonio Rago, Emanuele Albini, Pietro Baroni, Francesca Toni

This survey paper consolidates research on Explainable AI (XAI) that leverages computational argumentation frameworks, classifying existing methods based on explanation type, the AI model being explained, and the specific argumentation formalism employed. It also outlines a research roadmap for further development in this specialized field.

Key insights

  • Computational argumentation offers a robust, flexible, and theoretically grounded paradigm for generating explanations across diverse AI systems.
  • Argumentative explanations are applicable to various AI model types, including recommender systems, classifiers, decision-making systems, and logic programming tools.
  • Formal argumentation frameworks provide a powerful mechanism to represent complex reasoning for both inherently argumentative AI models and for providing post-hoc explanations of non-argumentative systems.

Summary by alphaXiv.

Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.