Heterogeneous graph neural networks for assumption-based argumentation
File(s)
Accepted version
Author(s)
Gehlot, Preesha
Rapberger, Anna
Russo, Fabrizio
Toni, Francesca
Type
Conference Paper
Abstract
Assumption-Based Argumentation (ABA) is a powerful
structured argumentation formalism, but exact computation of extensions under stable semantics is intractable for large frameworks. We present the first Graph Neural Network (GNN) approach to approximate credulous acceptance in ABA. To leverage GNNs, we model ABA frameworks via a dependency graph representation encoding assumptions, claims and rules as nodes, with heterogeneous edge labels distinguishing support, derive and attack relations. We propose two GNN architectures—ABAGCN and ABAGAT—that stack residual heterogeneous convolution or attention layers, respectively, to learn node embeddings. Our models are trained on the ICCMA 2023 benchmark, augmented with synthetic ABAFs, with hyperparameters optimised via Bayesian search. Empirically, both ABAGCN and ABAGAT outperform a state-of-the-art GNN baseline that we adapt from the abstract argumentation literature, achieving a node-level F1 score of up to 0.71 on the ICCMA instances. Finally, we develop a sound polynomial time extension-reconstruction algorithm driven by our predictor: it reconstructs stable extensions with F1 above 0.85 on small ABAFs and maintains an F1 of about 0.58 on large frameworks. Our
work opens new avenues for scalable approximate reasoning in structured argumentation.
structured argumentation formalism, but exact computation of extensions under stable semantics is intractable for large frameworks. We present the first Graph Neural Network (GNN) approach to approximate credulous acceptance in ABA. To leverage GNNs, we model ABA frameworks via a dependency graph representation encoding assumptions, claims and rules as nodes, with heterogeneous edge labels distinguishing support, derive and attack relations. We propose two GNN architectures—ABAGCN and ABAGAT—that stack residual heterogeneous convolution or attention layers, respectively, to learn node embeddings. Our models are trained on the ICCMA 2023 benchmark, augmented with synthetic ABAFs, with hyperparameters optimised via Bayesian search. Empirically, both ABAGCN and ABAGAT outperform a state-of-the-art GNN baseline that we adapt from the abstract argumentation literature, achieving a node-level F1 score of up to 0.71 on the ICCMA instances. Finally, we develop a sound polynomial time extension-reconstruction algorithm driven by our predictor: it reconstructs stable extensions with F1 above 0.85 on small ABAFs and maintains an F1 of about 0.58 on large frameworks. Our
work opens new avenues for scalable approximate reasoning in structured argumentation.
Date Issued
2026-03-14
Date Acceptance
2025-11-08
Citation
Proceedings of the AAAI Conference on Artificial Intelligence, 2026, 40 (23 AAAI-26 Technical Tracks 23), pp.19117-19125
ISSN
2374-3468
Publisher
Association for the Advancement of Artificial Intelligence
Start Page
19117
End Page
19125
Journal / Book Title
Proceedings of the AAAI Conference on Artificial Intelligence
Volume
40
Issue
23 AAAI-26 Technical Tracks 23
Copyright Statement
Copyright © 2026, Association for the Advancement of Artificial Intelligence.
Source
The 40th Annual AAAI Conference on Artificial Intelligence
Publication Status
Published
Start Date
2026-01-20
Finish Date
2026-01-27
Coverage Spatial
Singapore