As we know, Machine Learning is ubiquitous in our day to day lives. From product recommendations on Amazon, targeted advertising, and suggestions of what to watch, to funny Instagram filters. If something goes wrong with these, it probably won't ruin your life. Maybe you won't get that perfect selfie, or maybe companies will have to spend more on advertising. We need to be able to dissect our model, we will need to be able to understand and explain our model before it goes anywhere near a production system.

The study of computational models for argumentation is a vibrant area of artificial intelligence and, in particular, knowledge representation and reasoning research. Arguments most often have an intrinsic structure made explicit through derivations from more basic structures. Computational models for structured argumentation enable making the internal structure of arguments explicit. Assumption-based argumentation (ABA) is a central structured formalism for argumentation in AI. In this article, we make both algorithmic and complexity-theoretic advances in the study of ABA. In terms of algorithms, we propose a new approach to reasoning in a commonly studied fragment of ABA (namely the logic programming fragment) with and without preferences. While previous approaches to reasoning over ABA frameworks apply either specialized algorithms or translate ABA reasoning to reasoning over abstract argumentation frameworks, we develop a direct declarative approach to ABA reasoning by encoding ABA reasoning tasks in answer set programming. We show via an extensive empirical evaluation that our approach significantly improves on the empirical performance of current ABA reasoning systems. In terms of computational complexity, while the complexity of reasoning over ABA frameworks is well-understood, the complexity of reasoning in the ABA+ formalism integrating preferences into ABA is currently not fully established. Towards bridging this gap, our results suggest that the integration of preferential information into ABA via so-called reverse attacks results in increased problem complexity for several central argumentation semantics.

This volume contains revised versions of the papers selected for the second volume of the Online Handbook of Argumentation for AI (OHAAI). Previously, formal theories of argument and argument interaction have been proposed and studied, and this has led to the more recent study of computational models of argument. Argumentation, as a field within artificial intelligence (AI), is highly relevant for researchers interested in symbolic representations of knowledge and defeasible reasoning. The purpose of this handbook is to provide an open access and curated anthology for the argumentation research community. OHAAI is designed to serve as a research hub to keep track of the latest and upcoming PhD-driven research on the theory and application of argumentation in all areas related to AI.

NLP or natural language processing, human-like comprehension and explainable AI all sound like buzz words, but they are all incredibly important. Luca Scagliarini is chief product officer at Expert.AI and has been looking into NLP long before it was cool. The full interview is in this video, but here is some background. Watch the video for the full interview and to lear more about Luca Scagliarini's wealth of knowledge on various AI technologies and how they are being applied.

In many real-life situations that involve exchanges of arguments, individuals may differ on their assessment of which supports between the arguments are in fact justified, i.e., they put forward different support-relations. When confronted with such situations, we may wish to aggregate individuals' argumentation views on support-relations into a collective view, which is acceptable to the group. In this paper, we assume that under bipolar argumentation frameworks, individuals are equipped with a set of arguments and a set of attacks between arguments, but with possibly different support-relations. Using the methodology in social choice theory, we analyze what semantic properties of bipolar argumentation frameworks can be preserved by aggregation rules during the aggregation of support-relations.

Explaining the predictions of opaque machine learning algorithms is an important and challenging task, especially as complex models are increasingly used to assist in high-stakes decisions such as those arising in healthcare and finance. Most popular tools for post-hoc explainable artificial intelligence (XAI) are either insensitive to context (e.g., feature attributions) or difficult to summarize (e.g., counterfactuals). In this paper, I introduce \emph{rational Shapley values}, a novel XAI method that synthesizes and extends these seemingly incompatible approaches in a rigorous, flexible manner. I leverage tools from decision theory and causal modeling to formalize and implement a pragmatic approach that resolves a number of known challenges in XAI. By pairing the distribution of random variables with the appropriate reference class for a given explanation task, I illustrate through theory and experiments how user goals and knowledge can inform and constrain the solution set in an iterative fashion. The method compares favorably to state of the art XAI tools in a range of quantitative and qualitative comparisons.

Counterfactual explanations and adversarial examples have emerged as critical research areas for addressing the explainability and robustness goals of machine learning (ML). While counterfactual explanations were developed with the goal of providing recourse to individuals adversely impacted by algorithmic decisions, adversarial examples were designed to expose the vulnerabilities of ML models. While prior research has hinted at the commonalities between these frameworks, there has been little to no work on systematically exploring the connections between the literature on counterfactual explanations and adversarial examples. In this work, we make one of the first attempts at formalizing the connections between counterfactual explanations and adversarial examples. More specifically, we theoretically analyze salient counterfactual explanation and adversarial example generation methods, and highlight the conditions under which they behave similarly. Our analysis demonstrates that several popular counterfactual explanation and adversarial example generation methods such as the ones proposed by Wachter et. al. and Carlini and Wagner (with mean squared error loss), and C-CHVAE and natural adversarial examples by Zhao et. al. are equivalent. We also bound the distance between counterfactual explanations and adversarial examples generated by Wachter et. al. and DeepFool methods for linear models. Finally, we empirically validate our theoretical findings using extensive experimentation with synthetic and real world datasets.

Adversarial images highlight how vulnerable modern image classifiers are to perturbations outside of their training set. Human oversight might mitigate this weakness, but depends on humans understanding the AI well enough to predict when it is likely to make a mistake. In previous work we have found that humans tend to assume that the AI's decision process mirrors their own. Here we evaluate if methods from explainable AI can disrupt this assumption to help participants predict AI classifications for adversarial and standard images. We find that both saliency maps and examples facilitate catching AI errors, but their effects are not additive, and saliency maps are more effective than examples.

Explainable AI (XAI) methods are used in order to improve the interpretability of these complex "black box" models, thereby increasing transparency and enabling informed decision-making (Guidotti et al, 2018). Despite this, methods to assess the quality of explanations generated by such explainable methods are so far under-explored. In particular, functionallygrounded evaluation methods, which measure the inherent ability of explainable methods in a given situation, are often specific to a particular type of dataset or explainable method. A key measure of functionally-grounded explanation fitness is explanation fidelity, which assesses the correctness and completeness of the explanation with respect to the underlying black box predictive model (Zhou et al, 2021). Evaluations of fidelity in literature can generally be classified as one of the following: external fidelity evaluation, which assesses how well the prediction of the underlying model and the explanation agree, and internal fidelity, which assesses how well the explanation matches the decision-making processes of the underlying model (Messalas et al, 2019). While methods to evaluate external fidelity are relatively common in literature (Guidotti et al, 2019; Lakkaraju et al, 2016; Ming et al, 2019; Shankaranarayana and Runje, 2019), evaluation methods to evaluate internal fidelity using black box models are generally limited to text and image data, rather than tabular (Du et al, 2019; Fong and Vedaldi, 2017; Nguyen, 2018; Samek et al, 2017). In this paper, weproposeanovelevaluation method based onathree phase approach:(1) the creation of a fully transparent, inherently interpretable white box model, and evaluation of explanations against this model; (2) the usage of the white box as a proxy to refine and improve the evaluation of explanations generated by a black box model; and (3) test the fidelity of explanations for a black box model using the refined method from the second phase. The main contributions of this work are as follows: 1.

In recent research, human-understandable explanations of machine learning models have received a lot of attention. Often explanations are given in form of model simplifications or visualizations. However, as shown in cognitive science as well as in early AI research, concept understanding can also be improved by the alignment of a given instance for a concept with a similar counterexample. Contrasting a given instance with a structurally similar example which does not belong to the concept highlights what characteristics are necessary for concept membership. Such near misses have been proposed by Winston (1970) as efficient guidance for learning in relational domains. We introduce an explanation generation algorithm for relational concepts learned with Inductive Logic Programming (\textsc{GeNME}). The algorithm identifies near miss examples from a given set of instances and ranks these examples by their degree of closeness to a specific positive instance. A modified rule which covers the near miss but not the original instance is given as an explanation. We illustrate \textsc{GeNME} with the well known family domain consisting of kinship relations, the visual relational Winston arches domain and a real-world domain dealing with file management. We also present a psychological experiment comparing human preferences of rule-based, example-based, and near miss explanations in the family and the arches domains.