Research Papers

Sequential Nature of Recommender Systems Disrupts the Evaluation Process

Ali Shirali

Datasets are often generated in a sequential manner, where the previous samples and intermediate decisions or interventions affect subsequent samples. This is especially prominent in cases where there are significant human-AI interactions, such as in recommender systems. To characterize the importance of this relationship across samples, we propose to use adversarial attacks on popular evaluation processes. We present sequence-aware boosting attacks and provide a lower bound on the amount of extra information that can be exploited from a confidential test set solely based on the order of the observed data. We use real and synthetic data to test our methods and show that the evaluation process on the MovieLense-100k dataset can be affected by $\sim1\%$ which is important when considering the close competition. Codes are publicly available.

As we discussed in Part I of this topic, there is a clear desire to model and comprehend human behavior. Given the popular presupposition of human reasoning as the standard for learning and decision-making, there have been significant efforts and a growing trend in research to replicate these innate human abilities in artificial systems. In Part I, we discussed learning methods which generate a model of behavior from exploration of the system and feedback based on the exhibited behavior as well as topics relating to the use of or accounting for beliefs with respect to applicable skills or mental states of others. In this work, we will continue the discussion from the perspective of methods which focus on the assumed cognitive abilities, limitations, and biases demonstrated in human reasoning. We will arrange these topics as follows (i) methods such as cognitive architectures, cognitive heuristics, and related which demonstrate assumptions of limitations on cognitive resources and how that impacts decisions and (ii) methods which generate and utilize representations of bias or uncertainty to model human decision-making or the future outcomes of decisions.

The increased predictive power of machine learning models comes at the cost of increased complexity and loss of interpretability, particularly in comparison to parametric statistical models. This trade-off has led to the emergence of eXplainable AI (XAI) which provides methods, such as local explanations (LEs) and local variable attributions (LVAs), to shed light on how a model use predictors to arrive at a prediction. These provide a point estimate of the linear variable importance in the vicinity of a single observation. However, LVAs tend not to effectively handle association between predictors. To understand how the interaction between predictors affects the variable importance estimate, we can convert LVAs into linear projections and use the radial tour. This is also useful for learning how a model has made a mistake, or the effect of outliers, or the clustering of observations. The approach is illustrated with examples from categorical (penguin species, chocolate types) and quantitative (soccer/football salaries, house prices) response models. The methods are implemented in the R package cheem, available on CRAN.

In this meta-ethnography, we explore three different angles of ethical artificial intelligence (AI) design implementation including the philosophical ethical viewpoint, the technical perspective, and framing through a political lens. Our qualitative research includes a literature review that highlights the cross-referencing of these angles by discussing the value and drawbacks of contrastive top-down, bottom-up, and hybrid approaches previously published. The novel contribution to this framework is the political angle, which constitutes ethics in AI either being determined by corporations and governments and imposed through policies or law (coming from the top), or ethics being called for by the people (coming from the bottom), as well as top-down, bottom-up, and hybrid technicalities of how AI is developed within a moral construct and in consideration of its users, with expected and unexpected consequences and long-term impact in the world. There is a focus on reinforcement learning as an example of a bottom-up applied technical approach and AI ethics principles as a practical top-down approach. This investigation includes real-world case studies to impart a global perspective, as well as philosophical debate on the ethics of AI and theoretical future thought experimentation based on historical facts, current world circumstances, and possible ensuing realities.

AI-based systems have been used widely across various industries for different decisions ranging from operational decisions to tactical and strategic ones in low- and high-stakes contexts. Gradually the weaknesses and issues of these systems have been publicly reported including, ethical issues, biased decisions, unsafe outcomes, and unfair decisions, to name a few. Research has tended to optimize AI less has focused on its risk and unexpected negative consequences. Acknowledging this serious potential risks and scarcity of re-search I focus on unsafe outcomes of AI. Specifically, I explore this issue from a Human-AI interaction lens during AI deployment. It will be discussed how the interaction of individuals and AI during its deployment brings new concerns, which need a solid and holistic mitigation plan. It will be dis-cussed that only AI algorithms' safety is not enough to make its operation safe. The AI-based systems' end-users and their decision-making archetypes during collaboration with these systems should be considered during the AI risk management. Using some real-world scenarios, it will be highlighted that decision-making archetypes of users should be considered a design principle in AI-based systems.

Effective Representation to Capture Collaboration Behaviors between Explainer and User

Arjun Akula, Song-Chun Zhu

An explainable AI (XAI) model aims to provide transparency (in the form of justification, explanation, etc) for its predictions or actions made by it. Recently, there has been a lot of focus on building XAI models, especially to provide explanations for understanding and interpreting the predictions made by deep learning models. At UCLA, we propose a generic framework to interact with an XAI model in natural language.

Recent advances in artificial intelligence applied to biomedical text are opening exciting opportunities for improving pharmacovigilance activities currently burdened by the ever growing volumes of real world data. To fully realize these opportunities, existing regulatory guidance and industry best practices should be taken into consideration in order to increase the overall trustworthiness of the system and enable broader adoption. In this paper we present a case study on how to operationalize existing guidance for validated AI systems in pharmacovigilance focusing on the specific task of medical literature monitoring (MLM) of adverse events from the scientific literature. We describe an AI system designed with the goal of reducing effort in MLM activities built in close collaboration with subject matter experts and considering guidance for validated systems in pharmacovigilance and AI transparency. In particular we make use of public disclosures as a useful risk control measure to mitigate system misuse and earn user trust. In addition we present experimental results showing the system can significantly remove screening effort while maintaining high levels of recall (filtering 55% of irrelevant articles on average, for a target recall of 0.99 on suspected adverse articles) and provide a robust method for tuning the desired recall to suit a particular risk profile.

Characterizing Human Explanation Strategies to Inform the Design of Explainable AI for Building Damage Assessment

Donghoon Shin, Sachin Grover, Kenneth Holstein +1

Explainable AI (XAI) is a promising means of supporting human-AI collaborations for high-stakes visual detection tasks, such as damage detection tasks from satellite imageries, as fully-automated approaches are unlikely to be perfectly safe and reliable. However, most existing XAI techniques are not informed by the understandings of task-specific needs of humans for explanations. Thus, we took a first step toward understanding what forms of XAI humans require in damage detection tasks. We conducted an online crowdsourced study to understand how people explain their own assessments, when evaluating the severity of building damage based on satellite imagery. Through the study with 60 crowdworkers, we surfaced six major strategies that humans utilize to explain their visual damage assessments. We present implications of our findings for the design of XAI methods for such visual detection contexts, and discuss opportunities for future research.

Detecting and Mitigating Test-time Failure Risks via Model-agnostic Uncertainty Learning

Preethi Lahoti, Krishna P. Gummadi, Gerhard Weikum

Reliably predicting potential failure risks of machine learning (ML) systems when deployed with production data is a crucial aspect of trustworthy AI. This paper introduces Risk Advisor, a novel post-hoc meta-learner for estimating failure risks and predictive uncertainties of any already-trained black-box classification model. In addition to providing a risk score, the Risk Advisor decomposes the uncertainty estimates into aleatoric and epistemic uncertainty components, thus giving informative insights into the sources of uncertainty inducing the failures. Consequently, Risk Advisor can distinguish between failures caused by data variability, data shifts and model limitations and advise on mitigation actions (e.g., collecting more data to counter data shift). Extensive experiments on various families of black-box classification models and on real-world and synthetic datasets covering common ML failure scenarios show that the Risk Advisor reliably predicts deployment-time failure risks in all the scenarios, and outperforms strong baselines.

Deep neural networks, especially convolutional deep neural networks, are state-of-the-art methods to classify, segment or even generate images, movies, or sounds. However, these methods lack of a good semantic understanding of what happens internally. The question, why a COVID-19 detector has classified a stack of lung-ct images as positive, is sometimes more interesting than the overall specificity and sensitivity. Especially when human domain expert knowledge disagrees with the given output. This way, human domain experts could also be advised to reconsider their choice, regarding the information pointed out by the system. In addition, the deep learning model can be controlled, and a present dataset bias can be found. Currently, most explainable AI methods in the computer vision domain are purely used on image classification, where the images are ordinary images in the visible spectrum. As a result, there is no comparison on how the methods behave with multimodal image data, as well as most methods have not been investigated on how they behave when used for object detection. This work tries to close the gaps. Firstly, investigating three saliency map generator methods on how their maps differ across the different spectra. This is achieved via accurate and systematic training. Secondly, we examine how they behave when used for object detection. As a practical problem, we chose object detection in the infrared and visual spectrum for autonomous driving. The dataset used in this work is the Multispectral Object Detection Dataset, where each scene is available in the FIR, MIR and NIR as well as visual spectrum. The results show that there are differences between the infrared and visual activation maps. Further, an advanced training with both, the infrared and visual data not only improves the network's output, it also leads to more focused spots in the saliency maps.

The underlying hypothesis of knowledge-based explainable artificial intelligence is the data required for data-centric artificial intelligence agents (e.g., neural networks) are less diverse in contents than the data required to explain the decisions of such agents to humans. The idea is that a classifier can attain high accuracy using data that express a phenomenon from one perspective whereas the audience of explanations can entail multiple stakeholders and span diverse perspectives. We hence propose to use domain knowledge to complement the data used by agents. We formulate knowledge-based explainable artificial intelligence as a supervised data classification problem aligned with the CBR methodology. In this formulation, the inputs are case problems composed of both the inputs and outputs of the data-centric agent and case solutions, the outputs, are explanation categories obtained from domain knowledge and subject matter experts. This formulation does not typically lead to an accurate classification, preventing the selection of the correct explanation category. Knowledge-based explainable artificial intelligence extends the data in this formulation by adding features aligned with domain knowledge that can increase accuracy when selecting explanation categories.

Recently, the use of sound measures and metrics in Artificial Intelligence has become the subject of interest of academia, government, and industry. Efforts towards measuring different phenomena have gained traction in the AI community, as illustrated by the publication of several influential field reports and policy documents. These metrics are designed to help decision takers to inform themselves about the fast-moving and impacting influences of key advances in Artificial Intelligence in general and Machine Learning in particular. In this paper we propose to use such newfound capabilities of AI technologies to augment our AI measuring capabilities. We do so by training a model to classify publications related to ethical issues and concerns. In our methodology we use an expert, manually curated dataset as the training set and then evaluate a large set of research papers. Finally, we highlight the implications of AI metrics, in particular their contribution towards developing trustful and fair AI-based tools and technologies. Keywords: AI Ethics; AI Fairness; AI Measurement. Ethics in Computer Science.