As the use of AI in society grows, addressing emerging biases is essential to prevent systematic discrimination. Several bias detection methods have been proposed, but, with few exceptions, these tend to ignore transparency. Instead, interpretability and explainability are core requirements for algorithmic fairness, even more so than for other algorithmic solutions, given the human-oriented nature of fairness. We present ABIDE (Argumentative BIas detection by DEbate), a novel framework that structures bias detection transparently as debate, guided by an underlying argument graph as understood in (formal and computational) argumentation. The arguments are about the success chances of groups in local neighbourhoods and the significance of these neighbourhoods. We evaluate ABIDE experimentally and demonstrate its strengths in performance against an argumentative baseline.

Contestable AI requires that AI-driven decisions align with given preferences. Various types of argumentation frameworks have been shown to support forms of contestability. In this paper we focus on the little-studied Edge-Weighted Quantitative Bipolar Argumentation Frameworks (EW-QBAFs), where arguments have a base score as in QBAFs but attacks and supports (edges) are weighted. After generalising gradual semantics and properties thereof from QBAFs to EW-QBAFs, we introduce the contestability problem for EW-QBAFs, which asks how to modify edge weights to achieve a desired strength for a specific topic argument. To address this problem, we propose gradient-based relation attribution explanations (G-RAEs), which quantify the sensitivity of the topic argument's strength to changes in individual edge weights, thus providing interpretable guidance for weight adjustments towards contestability. Building on G-RAEs, we develop a heuristic algorithm that progressively adjusts the edge weights to attain the desired strength. We evaluate our approach experimentally on synthetic EW-QBAFs that simulate the structural characteristics of personalised recommender systems and multi-layer perceptrons, demonstrating that it can support contestability effectively.

We propose ProtoArgNet, a novel interpretable deep neural architecture for image classification in the spirit of prototypical-part-learning as found, e.g., in ProtoPNet. While earlier approaches associate every class with multiple prototypical-parts, ProtoArgNet uses super-prototypes that combine prototypical-parts into a unified class representation. This is done by combining local activations of prototypes in an MLP-like manner, enabling the localization of prototypes and learning (non-linear) spatial relationships among them. By leveraging a form of argumentation, ProtoArgNet is capable of providing both supporting (i.e. `this looks like that') and attacking (i.e. `this differs from that') explanations. We demonstrate on several datasets that ProtoArgNet outperforms state-of-the-art prototypical-part-learning approaches. Moreover, the argumentation component in ProtoArgNet is customisable to the user's cognitive requirements by a process of sparsification, which leads to more compact explanations compared to state-of-the-art approaches.

Yuqicheng Zhu, Nico Potyka, Jiarong Pan, Bo Xiong, Yunjie He, Evgeny Kharlamov, Steffen Staab. Proceedings of the 2025 Conference of the Nations of the Americas Chapter of the Association for Computational Linguistics: Human Language Technologies (Volume 1: Long Papers). 2025.

Retrieval-Augmented Generation (RAG) enhances large language models by incorporating external knowledge, yet suffers from critical limitations in high-stakes domains—namely, sensitivity to noisy or contradictory evidence and opaque, stochastic decision-making. We propose \textsc{ArgRAG}, an explainable, and contestable alternative that replaces black-box reasoning with structured inference using a Quantitative Bipolar Argumentation Framework (QBAF). \textsc{ArgRAG} constructs a QBAF from retrieved documents and performs deterministic reasoning under gradual semantics. This allows faithfully explanaining and contesting decisions. Evaluated on two fact verification benchmarks, PubHealth and RAGuard, \textsc{ArgRAG} achieves strong accuracy while significantly improving transparency.

Assumption-based Argumentation (ABA) is a well-known structured argumentation formalism, whereby arguments and attacks between them are drawn from rules, defeasible assumptions and their contraries. A common restriction imposed on ABA frameworks (ABAFs) is that they are flat, i.e. each of the defeasible assumptions can only be assumed, but not derived. While it is known that flat ABAFs can be translated into abstract argumentation frameworks (AFs) as proposed by Dung, no translation exists from general, possibly non-flat ABAFs into any kind of abstract argumentation formalism. In this paper, we close this gap and show that bipolar AFs (BAFs) can instantiate general ABAFs. To this end we develop suitable, novel BAF semantics which borrow from the notion of deductive support. We investigate basic properties of our BAFs, including computational complexity, and prove the desired relation to ABAFs under several semantics.

Counterfactual explanations shed light on the decisions of black-box models by explaining how an input can be altered to obtain a favourable decision from the model (e.g., when a loan application has been rejected). However, as noted recently, counterfactual explainers may lack robustness in the sense that a minor change in the input can cause a major change in the explanation. This can cause confusion on the user side and open the door for adversarial attacks. In this paper, we study some sources of non-robustness. While there are fundamental reasons for why an explainer that returns a single counterfactual cannot be robust in all instances, we show that some interesting robustness guarantees can be given by reporting multiple rather than a single counterfactual. Unfortunately, the number of counterfactuals that need to be reported for the theoretical guarantees to hold can be prohibitively large. We therefore propose an approximation algorithm that uses a diversity criterion to select a feasible number of most relevant explanations and study its robustness empirically. Our experiments indicate that our method improves the state-of-the-art in generating robust explanations, while maintaining other desirable properties and providing competitive computational performance.

Truth discovery networks evaluate the trustworthiness of sources (e.g., websites) and their claims (e.g., the severity of a virus). Intuitively, the more trustworthy the sources of a claim, the more believable the claim and vice versa. Singleton noted that bipolar abstract argumentation could be a natural way to reason about these networks. We explain how this idea can be implemented naturally by quantitative bipolar argumentation frameworks (QBAFs) that we call TD-QBAFs. While most applications of QBAFs result in a (nearly) acyclic structure, TD-QBAFs have bi-directional edges and can feature complex cycles. The stability (convergence behaviour) of QBAFs in cyclic graphs is currently not well understood. While pathological examples of divergent QBAFs have been constructed, the problems seemed unlikely to occur in practice. However, convergence problems seem to be the rule rather than the exception for TD-QBAFs. We demonstrate how common QBAF semantics can fail to converge for very simple TD-QBAFs and discuss some of the potential causes. While this shows limitations of existing semantics, we also discuss how some previously proposed ideas can be used to mitigate the problems and demonstrate their effectiveness empirically.

There is a growing interest in understanding arguments' strength in Quantitative Bipolar Argumentation Frameworks (QBAFs). Most existing studies focus on attribution-based methods that explain an argument's strength by assigning importance scores to other arguments but fail to explain how to change the current strength to a desired one. To solve this issue, we introduce counterfactual explanations for QBAFs. We discuss problem variants and propose an iterative algorithm named Counterfactual Explanations for Quantitative bipolar Argumentation frameworks (CE-QArg). CE-QArg can identify valid and cost-effective counterfactual explanations based on two core modules, polarity and priority, which help determine the updating direction and magnitude for each argument, respectively. We discuss some formal properties of our counterfactual explanations and empirically evaluate CE-QArg on randomly generated QBAFs.

Quantitative bipolar argumentation frameworks (QBAFs) have various applications in areas like product recommendation, review aggregation and explaining machine learning models. QBAF semantics assign a strength to every argument that is based on an a priori belief and the strength of its attackers and supporters. Intuitively, a QBAF semantics is open-minded when it is unbiased in the sense that a priori beliefs can be given up eventually when sufficient arguments to the contrary are presented. While this behaviour is desirable in many applications, existing open-minded semantics also have the property that even very weak arguments will eventually eliminate the a priori beliefs. In this paper, we will study notions of conservativeness that demand that the deviation from the a priori beliefs is bounded by the strength of pro and contra arguments. We will discuss compatibility and conflicts with existing properties and present two new semantics with interesting semantical guarantees. To do so, we will build up on the framework of modular semantics and prove some general relationships between functional and semantical properties that are useful to simplify the study of new modular semantics.

We present a principle-based analysis of contribution functions for quantitative bipolar argumentation graphs that quantify the contribution of one argument to another. The introduced principles formalise the intuitions underlying different contribution functions as well as expectations one would have regarding the behaviour of contribution functions in general. As none of the covered contribution functions satisfies all principles, our analysis can serve as a tool that enables the selection of the most suitable function based on the requirements of a given use case.

Random forests are decision tree ensembles that can be used to solve a variety of machine learning problems. However, as the number of trees and their individual size can be large, their decision making process is often incomprehensible. We show that their decision process can be naturally represented as an argumentation problem, which allows creating global explanations via argumentative reasoning. We generalize sufficient and necessary argumentative explanations using a Markov network encoding, discuss the relevance of these explanations and establish relationships to families of abductive explanations from the literature. As the complexity of the explanation problems is high, we present an efficient approximation algorithm with probabilistic approximation guarantees.

Argumentative explainable AI has been advocated by several in recent years, with an increasing interest on explaining the reasoning outcomes of Argumentation Frameworks (AFs). While there is a considerable body of research on qualitatively explaining the reasoning outcomes of AFs with debates/disputes/dialogues in the spirit of extension-based semantics, explaining the quantitative reasoning outcomes of AFs under gradual semantics has not received much attention, despite widespread use in applications. In this paper, we contribute to filling this gap by proposing a novel theory of Argument Attribution Explanations (AAEs) by incorporating the spirit of feature attribution from machine learning in the context of Quantitative Bipolar Argumentation Frameworks (QBAFs): whereas feature attribution is used to determine the influence of features towards outputs of machine learning models, AAEs are used to determine the influence of arguments towards topic arguments of interest. We study desirable properties of AAEs, including some new ones and some partially adapted from the literature to our setting. To demonstrate the applicability of our AAEs in practice, we conclude by carrying out two case studies in the scenarios of fake news detection and movie recommender systems.

Neural networks (NNs) have various applications in AI, but explaining their decisions remains challenging. Existing approaches often focus on explaining how changing individual inputs affects NNs’ outputs. However, an explanation that is consistent with the input-output behaviour of an NN is not necessarily faithful to the actual mechanics thereof. In this paper, we exploit relationships between multi-layer perceptrons (MLPs) and quantitative argumentation frameworks (QAFs) to create argumentative explanations for the mechanics of MLPs. Our SpArX method first sparsifies the MLP while maintaining as much of the original structure as possible. It then translates the sparse MLP into an equivalent QAF to shed light on the underlying decision process of the MLP, producing global and/or local explanations. We demonstrate experimentally that SpArX can give more faithful explanations than existing approaches, while simultaneously providing deeper insights into the actual reasoning process of MLPs.

Evidence from studies, such as in science or medicine, often corresponds to conditional probability statements. Furthermore, evidence can conflict, in particular when coming from multiple studies. Whilst it is natural to make sense of such evidence using arguments, there is a lack of a systematic formalism for representing and reasoning with conditional probability statements in computational argumentation. We address this shortcoming by providing a formalization of conditional probabilistic argumentation based on probabilistic conditional logic. We provide a semantics and a collection of comprehensible inference rules that give different insights into evidence. We show how arguments constructed from proofs and attacks between them can be analyzed as arguments graphs using dialectical semantics and via the epistemic approach to probabilistic argumentation. Our approach allows for a transparent and systematic way of handling uncertainty that often arises in evidence.

Graph Convolutional Networks (GCNs) are powerful frameworks for learning embeddings of graph-structured data. GCNs are traditionally studied through the lens of Euclidean geometry. Recent works find that non-Euclidean Riemannian manifolds provide specific inductive biases for embedding hierarchical or spherical data. However, they cannot align well with data of mixed graph topologies. We consider a larger class of pseudo-Riemannian manifolds that generalize hyperboloid and sphere. We develop new geodesic tools that allow for extending neural network operations into geodesically disconnected pseudo-Riemannian manifolds. As a consequence, we derive a pseudo-Riemannian GCN that models data in pseudo-Riemannian manifolds of constant nonzero curvature in the context of graph neural networks. Our method provides a geometric inductive bias that is sufficiently flexible to model mixed heterogeneous topologies like hierarchical graphs with cycles. We demonstrate the representational capabilities of this method by applying it to the tasks of graph reconstruction, node classification, and link prediction on a series of standard graphs with mixed topologies. Empirical results demonstrate that our method outperforms Riemannian counterparts when embedding graphs of complex topologies.

We show that an interesting class of feed-forward neural networks can be understood as quantitative argumentation frameworks. This connection creates a bridge between research in Formal Argumentation and Machine Learning. We generalize the semantics of feed-forward neural networks to acyclic graphs and study the resulting computational and semantical properties in argumentation graphs. As it turns out, the semantics gives stronger guarantees than existing semantics that have been tailor-made for the argumentation setting. From a machine-learning perspective, the connection does not seem immediately helpful. While it gives intuitive meaning to some feed-forward-neural networks, they remain difficult to understand due to their size and density. However, the connection seems helpful for combining background knowledge in form of sparse argumentation networks with dense neural networks that have been trained for complementary purposes and for learning the parameters of quantitative argumentation frameworks in an end-to-end fashion from data.

We explain how abstract argumentation problems can be encoded as Markov networks. From a computational perspective, this allows reducing argumentation tasks like finding labellings or deciding credulous and sceptical acceptance to probabilistic inference tasks in Markov networks. From a semantical perspective, the resulting probabilistic argumentation models are interesting in their own right. In particular, they satisfy several of the properties proposed for epistemic probabilistic argumentation by Hunter and Thimm. We also consider an extension to frameworks with deductive support and show that it maintains many of the interesting guarantees of both approaches .

Bipolar abstract argumentation frameworks allow modeling decision problems by defining pro and contra arguments and their relationships. In some popular bipolar frameworks, there is an inherent tendency to favor either attack or support relationships. However, for some applications, it seems sensible to treat attack and support equally. Roughly speaking, turning an attack edge into a support edge, should just invert its meaning. We look at a recently introduced bipolar argumentation semantics and two novel alternatives and discuss their semantical and computational properties. Interestingly, the two novel semantics correspond to stable semantics if no support relations are present and maintain the computational complexity of stable semantics in general bipolar frameworks.

Consolidation describes the operation of restoring consistency in an inconsistent knowledge base. Here we consider this problem in the context of probabilistic conditional logic, a language that focuses on probabilistic conditionals (if-then rules). If a knowledge base, i. e., a set of probabilistic conditionals, is inconsistent traditional model-based inference techniques are not applicable. In this paper, we develop an approach to repair such knowledge bases that relies on a generalized notion of a model of a knowledge base that extends to classically inconsistent knowledge bases. We define a generalized approach to reasoning under maximum entropy on these generalized models and use it to repair the knowledge base. This approach is founded on previous work on inconsistency measures and we show that it is well-defined, provides a unique solution, and satisfies other desirable properties.

Dealing with uncertainty that is inherently present in any medical domain, is one of the major challenges when designing a medical decision support system. We demonstrate how probabilistic logic can be used to design medical knowledge bases at the example of analysing clinical brain tumor data. We use MECoRe, a system implementing probabilistic conditional logic, to create a knowledge base BT that contains medical knowledge originating from both statistical data as well as from medical experts. Any incomplete or unspecified knowledge is completed by MECoRe in an information-theoretically optimal way by employing the principle of maximum entropy. BT is evaluated with respect to a series of queries regarding diagnosis and prognosis, using a real documented patient case.