Research Papers

Social Construction of XAI: Do We Need One Definition to Rule Them All?

Upol Ehsan, Mark O. Riedl

There is a growing frustration amongst researchers and developers in Explainable AI (XAI) around the lack of consensus around what is meant by 'explainability'. Do we need one definition of explainability to rule them all? In this paper, we argue why a singular definition of XAI is neither feasible nor desirable at this stage of XAI's development. We view XAI through the lenses of Social Construction of Technology (SCOT) to explicate how diverse stakeholders (relevant social groups) have different interpretations (interpretative flexibility) that shape the meaning of XAI. Forcing a standardization (closure) on the pluralistic interpretations too early can stifle innovation and lead to premature conclusions. We share how we can leverage the pluralism to make progress in XAI without having to wait for a definitional consensus.

Causality-Inspired Taxonomy for Explainable Artificial Intelligence

Pedro C. Neto, Tiago Gonçalves, João Ribeiro Pinto +4

As two sides of the same coin, causality and explainable artificial intelligence (xAI) were initially proposed and developed with different goals. However, the latter can only be complete when seen through the lens of the causality framework. As such, we propose a novel causality-inspired framework for xAI that creates an environment for the development of xAI approaches. To show its applicability, biometrics was used as case study. For this, we have analysed 81 research papers on a myriad of biometric modalities and different tasks. We have categorised each of these methods according to our novel xAI Ladder and discussed the future directions of the field.

With the increasing pervasiveness of algorithms across industry and government, a growing body of work has grappled with how to understand their societal impact and ethical implications. Various methods have been used at different stages of algorithm development to encourage researchers and designers to consider the potential societal impact of their research. An understudied yet promising area in this realm is using participatory foresight to anticipate these different societal impacts. We employ crowdsourcing as a means of participatory foresight to uncover four different types of impact areas based on a set of governmental algorithmic decision making tools: (1) perceived valence, (2) societal domains, (3) specific abstract impact types, and (4) ethical algorithm concerns. Our findings suggest that this method is effective at leveraging the cognitive diversity of the crowd to uncover a range of issues. We further analyze the complexities within the interaction of the impact areas identified to demonstrate how crowdsourcing can illuminate patterns around the connections between impacts. Ultimately this work establishes crowdsourcing as an effective means of anticipating algorithmic impact which complements other approaches towards assessing algorithms in society by leveraging participatory foresight and cognitive diversity.

With the increasing application of machine learning in high-stake decision-making problems, potential algorithmic bias towards people from certain social groups poses negative impacts on individuals and our society at large. In the real-world scenario, many such problems involve positive and unlabeled data such as medical diagnosis, criminal risk assessment and recommender systems. For instance, in medical diagnosis, only the diagnosed diseases will be recorded (positive) while others will not (unlabeled). Despite the large amount of existing work on fairness-aware machine learning in the (semi-)supervised and unsupervised settings, the fairness issue is largely under-explored in the aforementioned Positive and Unlabeled Learning (PUL) context, where it is usually more severe. In this paper, to alleviate this tension, we propose a fairness-aware PUL method named FairPUL. In particular, for binary classification over individuals from two populations, we aim to achieve similar true positive rates and false positive rates in both populations as our fairness metric. Based on the analysis of the optimal fair classifier for PUL, we design a model-agnostic post-processing framework, leveraging both the positive examples and unlabeled ones. Our framework is proven to be statistically consistent in terms of both the classification error and the fairness metric. Experiments on the synthetic and real-world data sets demonstrate that our framework outperforms state-of-the-art in both PUL and fair classification.

Deep learning methods exhibit outstanding performance in synthetic aperture radar (SAR) image interpretation tasks. However, these are black box models that limit the comprehension of their predictions. Therefore, to meet this challenge, we have utilized explainable artificial intelligence (XAI) methods for the SAR image classification task. Specifically, we trained state-of-the-art convolutional neural networks for each polarization format on OpenSARUrban dataset and then investigate eight explanation methods to analyze the predictions of the CNN classifiers of SAR images. These XAI methods are also evaluated qualitatively and quantitatively which shows that Occlusion achieves the most reliable interpretation performance in terms of Max-Sensitivity but with a low-resolution explanation heatmap. The explanation results provide some insights into the internal mechanism of black-box decisions for SAR image classification.

Autonomous systems are being developed and deployed in situations that may require some degree of ethical decision-making ability. As a result, research in machine ethics has proliferated in recent years. This work has included using moral dilemmas as validation mechanisms for implementing decision-making algorithms in ethically-loaded situations. Using trolley-style problems in the context of autonomous vehicles as a case study, I argue (1) that this is a misapplication of philosophical thought experiments because (2) it fails to appreciate the purpose of moral dilemmas, and (3) this has potentially catastrophic consequences; however, (4) there are uses of moral dilemmas in machine ethics that are appropriate and the novel situations that arise in a machine-learning context can shed some light on philosophical work in ethics.

The last decades have seen significant advancements in non-invasive neuroimaging technologies that have been increasingly adopted to examine human brain development. However, these improvements have not necessarily been followed by more sophisticated data analysis measures that are able to explain the mechanisms underlying functional brain development. For example, the shift from univariate (single area in the brain) to multivariate (multiple areas in brain) analysis paradigms is of significance as it allows investigations into the interactions between different brain regions. However, despite the potential of multivariate analysis to shed light on the interactions between developing brain regions, artificial intelligence (AI) techniques applied render the analysis non-explainable. The purpose of this paper is to understand the extent to which current state-of-the-art AI techniques can inform functional brain development. In addition, a review of which AI techniques are more likely to explain their learning based on the processes of brain development as defined by developmental cognitive neuroscience (DCN) frameworks is also undertaken. This work also proposes that eXplainable AI (XAI) may provide viable methods to investigate functional brain development as hypothesised by DCN frameworks.

The European Union (EU) through the High-Level Expert Group on Artificial Intelligence (AI-HLEG) and the General Data Protection Regulation (GDPR) has recently posed an interesting challenge to the eXplainable AI (XAI) community, by demanding a more user-centred approach to explain Automated Decision-Making systems (ADMs). Looking at the relevant literature, XAI is currently focused on producing explainable software and explanations that generally follow an approach we could term One-Size-Fits-All, that is unable to meet a requirement of centring on user needs. One of the causes of this limit is the belief that making things explainable alone is enough to have pragmatic explanations. Thus, insisting on a clear separation between explainabilty (something that can be explained) and explanations, we point to explanatorY AI (YAI) as an alternative and more powerful approach to win the AI-HLEG challenge. YAI builds over XAI with the goal to collect and organize explainable information, articulating it into something we called user-centred explanatory discourses. Through the use of explanatory discourses/narratives we represent the problem of generating explanations for Automated Decision-Making systems (ADMs) into the identification of an appropriate path over an explanatory space, allowing explainees to interactively explore it and produce the explanation best suited to their needs.

Challenging common interpretability assumptions in feature attribution explanations

Jonathan Dinu, Jeffrey Bigham, J. Zico Kolter

As machine learning and algorithmic decision making systems are increasingly being leveraged in high-stakes human-in-the-loop settings, there is a pressing need to understand the rationale of their predictions. Researchers have responded to this need with explainable AI (XAI), but often proclaim interpretability axiomatically without evaluation. When these systems are evaluated, they are often tested through offline simulations with proxy metrics of interpretability (such as model complexity). We empirically evaluate the veracity of three common interpretability assumptions through a large scale human-subjects experiment with a simple "placebo explanation" control. We find that feature attribution explanations provide marginal utility in our task for a human decision maker and in certain cases result in worse decisions due to cognitive and contextual confounders. This result challenges the assumed universal benefit of applying these methods and we hope this work will underscore the importance of human evaluation in XAI research. Supplemental materials -- including anonymized data from the experiment, code to replicate the study, an interactive demo of the experiment, and the models used in the analysis -- can be found at: https://doi.pizza/challenging-xai.

This study proposes an innovative explainable predictive quality analytics solution to facilitate data-driven decision-making for process planning in manufacturing by combining process mining, machine learning, and explainable artificial intelligence (XAI) methods. For this purpose, after integrating the top-floor and shop-floor data obtained from various enterprise information systems, a deep learning model was applied to predict the process outcomes. Since this study aims to operationalize the delivered predictive insights by embedding them into decision-making processes, it is essential to generate relevant explanations for domain experts. To this end, two complementary local post-hoc explanation approaches, Shapley values and Individual Conditional Expectation (ICE) plots are adopted, which are expected to enhance the decision-making capabilities by enabling experts to examine explanations from different perspectives. After assessing the predictive strength of the applied deep neural network with relevant binary classification evaluation measures, a discussion of the generated explanations is provided.

Test, Evaluation, Verification, and Validation (TEVV) for Artificial Intelligence (AI) is a challenge that threatens to limit the economic and societal rewards that AI researchers have devoted themselves to producing. A central task of TEVV for AI is estimating brittleness, where brittleness implies that the system functions well within some bounds and poorly outside of those bounds. This paper argues that neither of those criteria are certain of Deep Neural Networks. First, highly touted AI successes (eg. image classification and speech recognition) are orders of magnitude more failure-prone than are typically certified in critical systems even within design bounds (perfectly in-distribution sampling). Second, performance falls off only gradually as inputs become further Out-Of-Distribution (OOD). Enhanced emphasis is needed on designing systems that are resilient despite failure-prone AI components as well as on evaluating and improving OOD performance in order to get AI to where it can clear the challenging hurdles of TEVV and certification.

White-box Induction From SVM Models: Explainable AI with Logic Programming

Farhad Shakerin, Gopal Gupta

We focus on the problem of inducing logic programs that explain models learned by the support vector machine (SVM) algorithm. The top-down sequential covering inductive logic programming (ILP) algorithms (e.g., FOIL) apply hill-climbing search using heuristics from information theory. A major issue with this class of algorithms is getting stuck in a local optimum. In our new approach, however, the data-dependent hill-climbing search is replaced with a model-dependent search where a globally optimal SVM model is trained first, then the algorithm looks into support vectors as the most influential data points in the model, and induces a clause that would cover the support vector and points that are most similar to that support vector. Instead of defining a fixed hypothesis search space, our algorithm makes use of SHAP, an example-specific interpreter in explainable AI, to determine a relevant set of features. This approach yields an algorithm that captures SVM model's underlying logic and outperforms %GG: the FOIL algorithm --> other ILP algorithms other ILP algorithms in terms of the number of induced clauses and classification evaluation metrics. This paper is under consideration for publication in the journal of "Theory and practice of logic programming".