Explainable Artificial Intelligence (XAI)

By Michael Ridley. 2022

The field of explainable artificial intelligence (XAI) advances techniques, processes, and strategies that provide explanations for the predictions, recommendations, and decisions of opaque and complex machine learning systems. Increasingly academic libraries are providing library users with systems, services, and collections created and delivered by machine learning. Academic libraries should adopt XAI as a tool set to verify and validate these resources, and advocate for public policy regarding XAI that serves libraries, the academy, and the public interest.

Subfields: Safety, Robustness & Interpretability

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Safety, Robustness & Interpretability

Citations in the corpus

Explainable Artificial Intelligence (XAI): Concepts, Taxonomies, Opportunities and Challenges toward Responsible AI (2019, positive): “In order to avoid limiting the effectiveness of the current generation of AI systems, eXplainable AI (XAI) [7] proposes creating a suite of ML techniques that 1) produce more explainable models while maintaining a high level of learning performance (e.”
FairSight: Visual Analytics for Fairness in Decision Making (2019, neutral): “Machine learning and AI approaches have been recently criticized for the lack of capability in reasoning and diagnosing the logic behind the produced decisions [18].”
Trustworthy Artificial Intelligence: A Review (2022, background): “The Defense Advanced Research Project Agency (DARPA) [81] also launched a program known as Explainable Artificial Intelligence (XAI), whose motive was to make these AI systems explainable and trustworthy.”
Reasons, Values, Stakeholders: A Philosophical Framework for Explainable Artificial Intelligence (2021, background): “On this basis, several academic and industrial research programs have been put into operation to develop the so-called “explainable AI” systems: AI systems that are able to explain themselves (see, for example, [7, 8]).”
Explainable Artificial Intelligence: Objectives, Stakeholders, and Future Research Opportunities (2020, unknown): “and effectively manage the emerging generation of artificially intelligent partners” (Gunning, 2017).”
The impact of artificial intelligence on learner–instructor interaction in online learning (2021, background): “Explainability refers to the ability to offer human-understandable justifications for the AI’s output or procedures (Gunning, 2017).”
No Explainability without Accountability: An Empirical Study of Explanations and Feedback in Interactive ML (2020, unknown): “Ongoing government research programs [27], focused academic conferences,1 and recent legislation on the “right to explanation” [26] have also fueled a general push for ML system transparency.”
Explainable, trustworthy, and ethical machine learning for healthcare: A survey (2021, background): “DARPA1 organized a group of civilians and military researchers in 2017 intending to develop new methodologies for making ML models explainable [11].”
Explaining anomalies detected by autoencoders using Shapley Additive Explanations (2021, background): “Gunning [35] divided the explanation effectiveness into five categories: mental model, task performance, trust assessment, correctability, and user satisfaction.”
Principles and Practice of Explainable Machine Learning (2020, background): “She again tries various candidates and she finds out that Random Forests achieve the 5Note that this informal view encourages a notional plot of explainability versus accuracy, as is common in informal discussions on the challenge of XAI [32, 88].”
The explanation game: a formal framework for interpretable machine learning (2020, unknown): “, 2019) and military operations (Gunning, 2017), where user trust is essential and errors could be catastrophic.”
Expanding Explainability: Towards Social Transparency in AI systems (2021, background): “Although there is no established consensus on the complete set of factors that makes an AI system explainable, XAI work commonly shares the goal of making an AI system’s functioning or decisions easy to understand by people [7, 14, 27, 31, 34, 62, 70, 82].”
A Survey on the Explainability of Supervised Machine Learning (2020, unknown): “In May 2017, the DARPA launched the program XAI (Gunning, 2017) which aims at providing explainable and highly accurate models.”
Human-in-the-loop machine learning: a state of the art (2022, background): “A mind map of XAI can be found at Fig.”
Bayesian Workflow (2020, unknown): “…be difficult to understand, hence the need for exploratory model analysis (Unwin et al., 2003, Wickham, 2006) and explainable AI (Chen et al., 2018, Gunning, 2017, Rudin, 2018), which complements methods for evaluating, comparing, and averaging models using a range of interrelated approaches,…”
A Survey on Federated Learning Systems: Vision, Hype and Reality for Data Privacy and Protection (2019, background): “FL models is an open problem [62], [148].”