Publications
430 results found
Rago A, Russo F, Albini E, et al., 2022, Forging argumentative explanations from causal models, Proceedings of the 5th Workshop on Advances in Argumentation in Artificial Intelligence 2021 co-located with the 20th International Conference of the Italian Association for Artificial Intelligence (AIxIA 2021), Publisher: CEUR Workshop Proceedings, Pages: 1-15, ISSN: 1613-0073
We introduce a conceptualisation for generating argumentation frameworks (AFs) from causal models for the purpose of forging explanations for models' outputs. The conceptualisation is based on reinterpreting properties of semantics of AFs as explanation moulds, which are means for characterising argumentative relations. We demonstrate our methodology by reinterpreting the property of bi-variate reinforcement in bipolar AFs, showing how the extracted bipolar AFs may be used as relation-based explanations for the outputs of causal models.
Lertvittayakumjorn P, Choshen L, Shnarch E, et al., 2022, GrASP: A Library for Extracting and Exploring Human-Interpretable Textual Patterns., Publisher: European Language Resources Association, Pages: 6093-6103
Oksanen J, Majumder A, Saunack K, et al., 2022, A Graph-Based Method for Unsupervised Knowledge Discovery from Financial Texts, 13th International Conference on Language Resources and Evaluation (LREC), Publisher: EUROPEAN LANGUAGE RESOURCES ASSOC-ELRA, Pages: 5412-5417
Gaskell A, Miao Y, Toni F, et al., 2022, Logically Consistent Adversarial Attacks for Soft Theorem Provers., Publisher: ijcai.org, Pages: 4129-4135
, 2022, Computational Models of Argument - Proceedings of COMMA 2022, Cardiff, Wales, UK, 14-16 September 2022, Publisher: IOS Press
, 2022, 3rd ACM International Conference on AI in Finance, ICAIF 2022, New York, NY, USA, November 2-4, 2022, Publisher: ACM
Ward FR, Belardinelli F, Toni F, 2022, A causal perspective on AI deception in games., Publisher: CEUR-WS.org
Sukpanichnant P, Rago A, Lertvittayakumjorn P, et al., 2021, LRP-based argumentative explanations for neural networks, XAI.it 2021 - Italian Workshop on Explainable Artificial Intelligence, Pages: 71-84, ISSN: 1613-0073
In recent years, there have been many attempts to combine XAI with the field of symbolic AI in order to generate explanations for neural networks that are more interpretable and better align with human reasoning, with one prominent candidate for this synergy being the sub-field of computational argumentation. One method is to represent neural networks with quantitative bipolar argumentation frameworks (QBAFs) equipped with a particular semantics. The resulting QBAF can then be viewed as an explanation for the associated neural network. In this paper, we explore a novel LRP-based semantics under a new QBAF variant, namely neural QBAFs (nQBAFs). Since an nQBAF of a neural network is typically large, the nQBAF must be simplified before being used as an explanation. Our empirical evaluation indicates that the manner of this simplification is all important for the quality of the resulting explanation.
Kotonya N, Spooner T, Magazzeni D, et al., 2021, Graph reasoning with context-aware linearization for interpretable fact extraction and verification, FEVER 2021, Publisher: Association for Computational Linguistics, Pages: 21-30
This paper presents an end-to-end system for fact extraction and verification using textual and tabular evidence, the performance of which we demonstrate on the FEVEROUS dataset. We experiment with both a multi-task learning paradigm to jointly train a graph attention network for both the task of evidence extraction and veracity prediction, as well as a single objective graph model for solely learning veracity prediction and separate evidence extraction. In both instances, we employ a framework for per-cell linearization of tabular evidence, thus allowing us to treat evidence from tables as sequences. The templates we employ for linearizing tables capture the context as well as the content of table data. We furthermore provide a case study to show the interpretability our approach. Our best performing system achieves a FEVEROUS score of 0.23 and 53% label accuracy on the blind test data.
Albini E, Rago A, Baroni P, et al., 2021, Influence-driven explanations for bayesian network classifiers, PRICAI 2021, Publisher: Springer Verlag, Pages: 88-100, ISSN: 0302-9743
We propose a novel approach to buildinginfluence-driven ex-planations(IDXs) for (discrete) Bayesian network classifiers (BCs). IDXsfeature two main advantages wrt other commonly adopted explanationmethods. First, IDXs may be generated using the (causal) influences between intermediate, in addition to merely input and output, variables within BCs, thus providing adeep, rather than shallow, account of theBCs’ behaviour. Second, IDXs are generated according to a configurable set of properties, specifying which influences between variables count to-wards explanations. Our approach is thusflexible and can be tailored to the requirements of particular contexts or users. Leveraging on this flexibility, we propose novel IDX instances as well as IDX instances cap-turing existing approaches. We demonstrate IDXs’ capability to explainvarious forms of BCs, and assess the advantages of our proposed IDX instances with both theoretical and empirical analyses.
Rago A, Cocarascu O, Bechlivanidis C, et al., 2021, Argumentation as a framework for interactive explanations for recommendations, KR 2020, 17th International Conference on Principles of Knowledge Representation and Reasoning, Publisher: IJCAI, Pages: 805-815, ISSN: 2334-1033
As AI systems become ever more intertwined in our personallives, the way in which they explain themselves to and inter-act with humans is an increasingly critical research area. Theexplanation of recommendations is, thus a pivotal function-ality in a user’s experience of a recommender system (RS),providing the possibility of enhancing many of its desirablefeatures in addition to itseffectiveness(accuracy wrt users’preferences). For an RS that we prove empirically is effective,we show how argumentative abstractions underpinning rec-ommendations can provide the structural scaffolding for (dif-ferent types of) interactive explanations (IEs), i.e. explana-tions empowering interactions with users. We prove formallythat these IEs empower feedback mechanisms that guaranteethat recommendations will improve with time, hence render-ing the RSscrutable. Finally, we prove experimentally thatthe various forms of IE (tabular, textual and conversational)inducetrustin the recommendations and provide a high de-gree oftransparencyin the RS’s functionality.
Cyras K, Rago A, Emanuele A, et al., 2021, Argumentative XAI: a survey, The 30th International Joint Conference on Artificial Intelligence (IJCAI-21), Publisher: International Joint Conferences on Artificial Intelligence, Pages: 4392-4399
Explainable AI (XAI) has been investigated for decades and, together with AI itself, has witnessed unprecedented growth in recent years. Among various approaches to XAI, argumentative models have been advocated in both the AI and social science literature, as their dialectical nature appears to match some basic desirable features of the explanation activity. In this survey we overview XAI approaches built using methods from the field of computational argumentation, leveraging its wide array of reasoning abstractions and explanation delivery methods. We overview the literature focusing on different types of explanation (intrinsic and post-hoc), different models with which argumentation-based explanations are deployed, different forms of delivery, and different argumentation frameworks they use. We also lay out a roadmap for future work.
Cocarascu O, Cyras K, Rago A, et al., 2021, Mining property-driven graphical explanations for data-centric AI from argumentation frameworks, Human-Like Machine Intelligence, Pages: 93-113, ISBN: 9780198862536
- Cite
- Citations: 1
Zylberajch H, Lertvittayakumjorn P, Toni F, 2021, HILDIF: interactive debugging of NLI models using influence functions, 1st Workshop on Interactive Learning for Natural Language Processing (InterNLP), Publisher: ASSOC COMPUTATIONAL LINGUISTICS-ACL, Pages: 1-6
Biases and artifacts in training data can cause unwelcome behavior in text classifiers (such as shallow pattern matching), leading to lack of generalizability. One solution to this problem is to include users in the loop and leverage their feedback to improve models. We propose a novel explanatory debugging pipeline called HILDIF, enabling humans to improve deep text classifiers using influence functions as an explanation method. We experiment on the Natural Language Inference (NLI) task, showing that HILDIF can effectively alleviate artifact problems in fine-tuned BERT models and result in increased model generalizability.
Albini E, Baroni P, Rago A, et al., 2021, Interpreting and explaining pagerank through argumentation semantics, Intelligenza Artificiale, Vol: 15, Pages: 17-34, ISSN: 1724-8035
In this paper we show how re-interpreting PageRank as an argumentation semantics for a bipolar argumentation framework empowers its explainability. After showing that PageRank, naively re-interpreted as an argumentation semantics for support frameworks, fails to satisfy some generally desirable properties, we propose a novel approach able to reconstruct PageRank as a gradual semantics of a suitably defined bipolar argumentation framework, while satisfying these properties. We then show how the theoretical advantages afforded by this approach also enjoy an enhanced explanatory power: we propose several types of argument-based explanations for PageRank, each of which focuses on different aspects of the algorithm and uncovers information useful for the comprehension of its results.
Paulino-Passos G, Toni F, 2021, Monotonicity and Noise-Tolerance in Case-Based Reasoning with Abstract Argumentation (with Appendix)
Recently, abstract argumentation-based models of case-based reasoning($AA{\text -} CBR$ in short) have been proposed, originally inspired by thelegal domain, but also applicable as classifiers in different scenarios.However, the formal properties of $AA{\text -} CBR$ as a reasoning systemremain largely unexplored. In this paper, we focus on analysing thenon-monotonicity properties of a regular version of $AA{\text -} CBR$ (that wecall $AA{\text -} CBR_{\succeq}$). Specifically, we prove that $AA{\text -}CBR_{\succeq}$ is not cautiously monotonic, a property frequently considereddesirable in the literature. We then define a variation of $AA{\text -}CBR_{\succeq}$ which is cautiously monotonic. Further, we prove that suchvariation is equivalent to using $AA{\text -} CBR_{\succeq}$ with a restrictedcasebase consisting of all "surprising" and "sufficient" cases in the originalcasebase. As a by-product, we prove that this variation of $AA{\text -}CBR_{\succeq}$ is cumulative, rationally monotonic, and empowers a principledtreatment of noise in "incoherent" casebases. Finally, we illustrate $AA{\text-} CBR$ and cautious monotonicity questions on a case study on the U.S. TradeSecrets domain, a legal casebase.
Rago A, Cocarascu O, Bechlivanidis C, et al., 2021, Argumentative explanations for interactive recommendations, Artificial Intelligence, Vol: 296, Pages: 1-22, ISSN: 0004-3702
A significant challenge for recommender systems (RSs), and in fact for AI systems in general, is the systematic definition of explanations for outputs in such a way that both the explanations and the systems themselves are able to adapt to their human users' needs. In this paper we propose an RS hosting a vast repertoire of explanations, which are customisable to users in their content and format, and thus able to adapt to users' explanatory requirements, while being reasonably effective (proven empirically). Our RS is built on a graphical chassis, allowing the extraction of argumentation scaffolding, from which diverse and varied argumentative explanations for recommendations can be obtained. These recommendations are interactive because they can be questioned by users and they support adaptive feedback mechanisms designed to allow the RS to self-improve (proven theoretically). Finally, we undertake user studies in which we vary the characteristics of the argumentative explanations, showing users' general preferences for more information, but also that their tastes are diverse, thus highlighting the need for our adaptable RS.
Cyras K, Oliveira T, Karamlou M, et al., 2021, Assumption-based argumentation with preferences and goals for patient-centric reasoning with interacting clinical guidelines, Argument and Computation, Vol: 12, Pages: 149-189, ISSN: 1946-2166
A paramount, yet unresolved issue in personalised medicine is that of automated reasoning with clinical guidelines in multimorbidity settings. This entails enabling machines to use computerised generic clinical guideline recommendations and patient-specific information to yield patient-tailored recommendations where interactions arising due to multimorbidities are resolved. This problem is further complicated by patient management desiderata, in particular the need to account for patient-centric goals as well as preferences of various parties involved. We propose to solve this problem of automated reasoning with interacting guideline recommendations in the context of a given patient by means of computational argumentation. In particular, we advance a structured argumentation formalism ABA+G (short for Assumption-Based Argumentation with Preferences (ABA+) and Goals) for integrating and reasoning with information about recommendations, interactions, patient’s state, preferences and prioritised goals. ABA+G combines assumption-based reasoning with preferences and goal-driven selection among reasoning outcomes. Specifically, we assume defeasible applicability of guideline recommendations with the general goal of patient well-being, resolve interactions (conflicts and otherwise undesirable situations) among recommendations based on the state and preferences of the patient, and employ patient-centered goals to suggest interaction-resolving, goal-importance maximising and preference-adhering recommendations. We use a well-established Transition-based Medical Recommendation model for representing guideline recommendations and identifying interactions thereof, and map the components in question, together with the given patient’s state, prioritised goals, and preferences over actions, to ABA+G for automated reasoning. In this, we follow principles of patient management and establish corresponding theoretical properties as well as illustrate our approach in realis
Dejl A, He P, Mangal P, et al., 2021, Argflow: A toolkit for deep argumentative explanations for neural networks, AAMAS, Pages: 1749-1751, ISSN: 1548-8403
In recent years, machine learning (ML) models have been successfully applied in a variety of real-world applications. However, they are often complex and incomprehensible to human users. This can decrease trust in their outputs and render their usage in critical settings ethically problematic. As a result, several methods for explaining such ML models have been proposed recently, in particular for black-box models such as deep neural networks (NNs). Nevertheless, these methods predominantly explain outputs in terms of inputs, disregarding the inner workings of the ML model computing those outputs. We present Argflow, a toolkit enabling the generation of a variety of 'deep' argumentative explanations (DAXs) for outputs of NNs on classification tasks.
Cyras K, Heinrich Q, Toni F, 2021, Computational complexity of flat and generic assumption-based argumentation, with and without probabilities, Artificial Intelligence, Vol: 293, Pages: 1-36, ISSN: 0004-3702
Reasoning with probabilistic information has recently attracted considerable attention in argumentation, and formalisms of Probabilistic Abstract Argumentation (PAA), Probabilistic Bipolar Argumentation (PBA) and Probabilistic Structured Argumentation (PSA) have been proposed. These foundational advances have been complemented with investigations on the complexity of some approaches to PAA and PBA, but not to PSA. We study the complexity of an existing form of PSA, namely Probabilistic Assumption-Based Argumentation (PABA), a powerful, implemented formalism which subsumes several forms of PAA and other forms of PSA. Specifically, we establish membership (general upper bounds) and completeness (instantiated lower bounds) of reasoning in PABA for the class FP#P (of functions with a #P-oracle for counting the solutions of an NP problem) with respect to newly introduced probabilistic verification, credulous and sceptical acceptance function problems under several ABA semantics. As a by-product necessary to establish PABA complexity results, we provide a comprehensive picture of the ABA complexity landscape (for both flat and generic, possibly non-flat ABA) for the classical decision problems of verification, existence, credulous and sceptical acceptance under those ABA semantics.
Rago A, Albini E, Baroni P, et al., 2021, Influence-driven explanations for bayesian network classifiers, Publisher: arXiv
One of the most pressing issues in AI in recent years has been the need toaddress the lack of explainability of many of its models. We focus onexplanations for discrete Bayesian network classifiers (BCs), targeting greatertransparency of their inner workings by including intermediate variables inexplanations, rather than just the input and output variables as is standardpractice. The proposed influence-driven explanations (IDXs) for BCs aresystematically generated using the causal relationships between variableswithin the BC, called influences, which are then categorised by logicalrequirements, called relation properties, according to their behaviour. Theserelation properties both provide guarantees beyond heuristic explanationmethods and allow the information underpinning an explanation to be tailored toa particular context's and user's requirements, e.g., IDXs may be dialecticalor counterfactual. We demonstrate IDXs' capability to explain various forms ofBCs, e.g., naive or multi-label, binary or categorical, and also integraterecent approaches to explanations for BCs from the literature. We evaluate IDXswith theoretical and empirical analyses, demonstrating their considerableadvantages when compared with existing explanation methods.
Albini E, Lertvittayakumjorn P, Rago A, et al., 2021, DAX: deep argumentative eXplanation for neural networks, Publisher: arXiv
Despite the rapid growth in attention on eXplainable AI (XAI) of late,explanations in the literature provide little insight into the actualfunctioning of Neural Networks (NNs), significantly limiting theirtransparency. We propose a methodology for explaining NNs, providingtransparency about their inner workings, by utilising computationalargumentation (a form of symbolic AI offering reasoning abstractions for avariety of settings where opinions matter) as the scaffolding underpinning DeepArgumentative eXplanations (DAXs). We define three DAX instantiations (forvarious neural architectures and tasks) and evaluate them empirically in termsof stability, computational cost, and importance of depth. We also conducthuman experiments with DAXs for text classification models, indicating thatthey are comprehensible to humans and align with their judgement, while alsobeing competitive, in terms of user acceptance, with existing approaches to XAIthat also have an argumentative spirit.
Lertvittayakumjorn P, Toni F, 2021, Explanation-based human debugging of nlp models: a survey, Transactions of the Association for Computational Linguistics, Vol: 9, Pages: 1508-1528, ISSN: 2307-387X
Debugging a machine learning model is hard since the bug usually involves the training data and the learning process. This becomes even harder for an opaque deep learning model if we have no clue about how the model actually works. In this survey, we review papers that exploit explanations to enable humans to give feedback and debug NLP models. We call this problem explanation-based human debugging (EBHD). In particular, we categorize and discuss existing work along three dimensions of EBHD (the bug context, the workflow, and the experimental setting), compile findings on how EBHD components affect the feedback providers, and highlight open problems that could be future research directions.
Santos M, Toni F, 2021, Artificial intelligence-empowered technology in the home, HOME IN THE DIGITAL AGE, Editors: Argandona, Malala, Peatfield, Publisher: ROUTLEDGE, Pages: 38-55, ISBN: 978-0-367-53017-4
- Author Web Link
- Cite
- Citations: 1
Paulino-Passos G, Toni F, 2021, Monotonicity and Noise-Tolerance in Case-Based Reasoning with Abstract Argumentation, Pages: 508-518
Albini E, Rago A, Baroni P, et al., 2021, Influence-Driven Explanations for Bayesian Network Classifiers., Publisher: Springer, Pages: 88-100
Cyras K, Rago A, Albini E, et al., 2021, Argumentative XAI: A Survey., Publisher: ijcai.org, Pages: 4392-4399
Lauren S, Belardinelli F, Toni F, 2020, Aggregating Bipolar Opinions, 20th International Conference on Autonomous Agents and Multiagent Systems
Kotonya N, Toni F, 2020, Explainable Automated Fact-Checking: A Survey, Barcelona. Spain, 28th International Conference on Computational Linguistics (COLING 2020), Publisher: International Committee on Computational Linguistics, Pages: 5430-5443
A number of exciting advances have been made in automated fact-checkingthanks to increasingly larger datasets and more powerful systems, leading toimprovements in the complexity of claims which can be accurately fact-checked.However, despite these advances, there are still desirable functionalitiesmissing from the fact-checking pipeline. In this survey, we focus on theexplanation functionality -- that is fact-checking systems providing reasonsfor their predictions. We summarize existing methods for explaining thepredictions of fact-checking systems and we explore trends in this topic.Further, we consider what makes for good explanations in this specific domainthrough a comparative analysis of existing fact-checking explanations againstsome desirable properties. Finally, we propose further research directions forgenerating fact-checking explanations, and describe how these may lead toimprovements in the research area.v
Kotonya N, Toni F, 2020, Explainable Automated Fact-Checking for Public Health Claims, 2020 Conference on Empirical Methods in Natural Language Processing (EMNLP(1) 2020), Publisher: ACL, Pages: 7740-7754
Fact-checking is the task of verifying the veracity of claims by assessing their assertions against credible evidence. The vast major-ity of fact-checking studies focus exclusively on political claims. Very little research explores fact-checking for other topics, specifically subject matters for which expertise is required. We present the first study of explainable fact-checking for claims which require specific expertise. For our case study we choose the setting of public health. To support this case study we construct a new datasetPUBHEALTHof 11.8K claims accompanied by journalist crafted, gold standard explanations(i.e., judgments) to support the fact-check la-bels for claims1. We explore two tasks: veracity prediction and explanation generation. We also define and evaluate, with humans and computationally, three coherence properties of explanation quality. Our results indicate that,by training on in-domain data, gains can be made in explainable, automated fact-checking for claims which require specific expertise.
This data is extracted from the Web of Science and reproduced under a licence from Thomson Reuters. You may not copy or re-distribute this data in whole or in part without the written consent of the Science business of Thomson Reuters.