Explanation in artificial intelligence: insights from the social sciences

DOI10.1016/J.ARTINT.2018.07.007zbMATH Open1478.68274arXiv1706.07269OpenAlexW2963095307WikidataQ102363022 ScholiaQ102363022MaRDI QIDQ2321252

Tim Miller

Publication date: 28 August 2019

Published in: Artificial Intelligence (Search for Journal in Brave)

Abstract: There has been a recent resurgence in the area of explainable artificial intelligence as researchers and practitioners seek to make their algorithms more understandable. Much of this research is focused on explicitly explaining decisions or actions to a human observer, and it should not be controversial to say that looking at how humans explain to each other can serve as a useful starting point for explanation in artificial intelligence. However, it is fair to say that most work in explainable artificial intelligence uses only the researchers' intuition of what constitutes a `good' explanation. There exists vast and valuable bodies of research in philosophy, psychology, and cognitive science of how people define, generate, select, evaluate, and present explanations, which argues that people employ certain cognitive biases and social expectations towards the explanation process. This paper argues that the field of explainable artificial intelligence should build on this existing research, and reviews relevant papers from philosophy, cognitive psychology/science, and social psychology, which study these topics. It draws out some important findings, and discusses ways that these can be infused with work on explainable artificial intelligence.

Full work available at URL: https://arxiv.org/abs/1706.07269

zbMATH Keywords

interpretabilitytransparencyexplanationexplainabilityexplainable AI

Mathematics Subject Classification ID

General topics in artificial intelligence (68T01)

Cites Work

• A Short Introduction to Computational Social Choice
• A theory of diagnosis from first principles
• Causes and Explanations: A Structural-Model Approach. Part I: Causes
• Causes and Explanations: A Structural-Model Approach. Part II: Explanations
• Complexity results for structure-based causality.
• Abductive Inference
• Causes and explanations in the structural-model approach: Tractable cases
• Conditional logic of actions and causation
• How to explain individual classification decisions
• Title not available (Why is that?)
• Title not available (Why is that?)
• Title not available (Why is that?)
• Title not available (Why is that?)
• Title not available (Why is that?)
• Title not available (Why is that?)
• Title not available (Why is that?)
• Title not available (Why is that?)
• Title not available (Why is that?)
• Title not available (Why is that?)

Cited In (90)

• Counterfactual state explanations for reinforcement learning agents via generative deep learning
• On computing probabilistic abductive explanations
• Mathematical optimization in classification and regression trees
• Counterfactuals as modal conditionals, and their probability
• Title not available (Why is that?)
• A Comprehensive Framework for Learning Declarative Action Models
• Using analogical proportions for explanations
• Probabilistic causes in Markov chains
• Defining formal explanation in classical logic by substructural derivability
• Explainable Deep Learning: A Field Guide for the Uninitiated
• Explaining black-box classifiers: properties and functions
• A Survey on the Explainability of Supervised Machine Learning
• A local method for identifying causal relations under Markov equivalence
• Relation between prognostics predictor evaluation metrics and local interpretability SHAP values
• ASP and subset minimality: enumeration, cautious reasoning and MUSes
• Learning Optimal Decision Sets and Lists with SAT
• Argumentative explanations for interactive recommendations
• Paracoherent answer set computation
• Why bad coffee? Explaining BDI agent behaviour with valuings
• The spherical \(k\)-means++ algorithm via local search
• Editable machine learning models? A rule-based framework for user studies of explainability
• Comments on ``Data science, big data and statistics
• On Tackling Explanation Redundancy in Decision Trees
• Witnesses for Answer Sets of Logic Programs
• Explainable acceptance in probabilistic and incomplete abstract argumentation frameworks
• Non-monotonic explanation functions
• Necessary and sufficient explanations for argumentation-based conclusions
• Persuasive contrastive explanations for Bayesian networks
• Tractability of explaining classifier decisions
• Explanation in AI and law: past, present and future
• Model Uncertainty and Correctability for Directed Graphical Models
• Title not available (Why is that?)
• Generating contrastive explanations for inductive logic programming based on a near miss approach
• Efficiently explaining CSPs with unsatisfiable subset optimization
• Beneficial and harmful explanatory machine learning
• Heterogeneous causal effects with imperfect compliance: a Bayesian machine learning approach
• Interpreting deep learning models with marginal attribution by conditioning on quantiles
• Model transparency and interpretability: survey and application to the insurance industry
• Local and global explanations of agent behavior: integrating strategy summaries with saliency maps
• Knowledge graphs as tools for explainable machine learning: a survey
• The quest of parsimonious XAI: a human-agent architecture for explanation formulation
• SAT-based rigorous explanations for decision lists
• Some thoughts on knowledge-enhanced machine learning
• Logic explained networks
• Toward an explainable machine learning model for claim frequency: a use case in car insurance pricing with telematics data
• Title not available (Why is that?)
• On cognitive preferences and the plausibility of rule-based models
• Formal Methods in FCA and Big Data
• A machine learning approach to differentiate between COVID-19 and influenza infection using synthetic infection and immune response data
• Detecting correlations and triangular arbitrage opportunities in the Forex by means of multifractal detrended cross-correlations analysis
• The explanation game: a formal framework for interpretable machine learning
• Story embedding: learning distributed representations of stories based on character networks
• A maximum-margin multisphere approach for binary multiple instance learning
• Exploiting Game Theory for Analysing Justifications
• The effects of explanations on automation bias
• Critical observations in model-based diagnosis
• Interpretable generalized additive neural networks
• What makes accidents severe! Explainable analytics framework with parameter optimization
• Explaining answers generated by knowledge graph embeddings
• Learning decision catalogues for situated decision making: the case of scoring systems
• On the failings of Shapley values for explainability
• Questionable stepwise explanations for a robust additive preference model
• Synergies between machine learning and reasoning -- an introduction by the Kay R. Amel group
• Synthesizing explainable counterfactual policies for algorithmic recourse with program synthesis
• Modifications of the Miller definition of contrastive (counterfactual) explanations
• A fast iterative PDE-based algorithm for feedback controls of nonsmooth mean-field control problems
• A framework for inherently interpretable optimization models
• Efficient search for relevance explanations using MAP-independence in Bayesian networks
• Providing personalized explanations: a conversational approach
• Objective-based counterfactual explanations for linear discrete optimization
• On the local coordination of fuzzy valuations
• Some models are useful, but how do we know which ones? Towards a unified Bayesian model taxonomy
• A \(k\)-additive Choquet integral-based approach to approximate the SHAP values for local interpretability in machine learning
• A general framework for personalising post hoc explanations through user knowledge integration
• Relative sparsity for medical decision problems
• Explainable subgradient tree boosting for prescriptive analytics in operations management
• Disagreement amongst counterfactual explanations: how transparency can be misleading
• Explainable and interpretable machine learning and data mining
• The Jiminy advisor: moral agreements among stakeholders based on norms and argumentation
• A logic of ``black box classifier systems
• Is there a role for statistics in artificial intelligence?
• A new model for counterfactual analysis for functional data
• On the (complete) reasons behind decisions
• Certified logic-based explainable AI -- the case of monotonic classifiers
• Feature necessity \& relevancy in ML classifier explanations
• Explanation-friendly query answering under uncertainty
• A new class of explanations for classifiers with non-binary features
• Contrastive explanations for answer-set programs
• Declarative reasoning on explanations using constraint logic programming
• Minimality, necessity and sufficiency for argumentation and explanation

This page was built for publication: Explanation in artificial intelligence: insights from the social sciences

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q2321252)