Inhalt

•  
   Preface
•  
   Contents
•  
   Contributors
•  
   Acronyms
•  
  1 Introducing Social Explainable AI
•  
   1.1 Societal Challenge and XAI
•  
   1.2 Challenges in XAI Development
•  
   1.3 Human Factors for XAI Development
•  
   1.4 Social Factors in XAI Development
•  
   1.5 The Focus on Interaction and a Definition of Social XAI
•  
   1.6 Addressees of the Book
•  
   1.7 Structure of the Book
•  
   References
•  
   2 Scenarios of Social Explainable AI in Practice
•  
   2.1 Assumptions for the Scenarios
•  
   2.2 Scenario 1: Explaining the Choice of a Car to Explainees with Different Backgrounds
•  
   2.3 Scenario 2: Decision Support System for Choosing a Car
•  
   2.4 Scenario 3: A Decision Support System for Medical Interventions
•  
   2.5 Scenario 4: Explaining in a Human–Human Tutoring Situation
•  
   2.6 Scenario 5: The Cognitive and Interactive Robots
•  
   2.7 Scenario 6: A Robot Teaching and Evaluating the Cooking
•  
   2.8 Elements of Social XAI Identified in the Scenarios
•  
   References
•  
   3 Components of an Explanation for Co-constructive sXAI
•  
   3.1 How Does This Chapter Relate to XAI?
•  
   3.2 The Role of Explainer and Explainee
•  
   3.3 Explanandum
•  
   3.4 Explanans
•  
   3.5 Operationalizations of the Explanation Components
•  
   3.6 How Does This Chapter Inspire Further Directions of XAI?
•  
   3.7 Rapid Access to the Content of This Chapter
•  
   References
•  
  4 Context for Explanations
•  
   4.1 How Does This Chapter Relate to XAI?
•  
   4.2 Making Sense of a Situation: Context in Humans
•  
   4.3 Situation-Aware XAI Systems: Context in AI
•  
   4.4 (Incrementally) Constructed Context: Semantic/Cognitive Interpretation of the Explanation Situation
•  
   4.5 Operationalization of Context
•  
   4.6 How Does This Chapter Inspire Further Directions of XAI?
•  
   4.7 Rapid Access to the Content of This Chapter
•  
   References
•  
   Part I Patterns
•  
   5 Practices: How to Establish an Explaining Practice
•  
   5.1 How Does This Chapter Relate to XAI?
•  
   5.2 A (Communicative) Practice
•  
   5.2.1 Multimodal Communicative Practice
•  
   5.2.2 How to Learn a Practice
•  
   5.3 An Explanation Practice
•  
   5.4 AI Systems and Practices
•  
   5.5 Operationalizations of Practices
•  
   5.6 How Does This Chapter Inspire Further Sirections of XAI?
•  
   5.7 Rapid Access to the Content of This Chapter
•  
   References
•  
  6 Explanation Goals
•  
   6.1 How Does This Chapter Relate to XAI?
•  
   6.2 Goals in Human Actions
•  
   6.3 Goals in AI
•  
   6.4 Goals in Collaborative Human–AI System Actions
•  
   6.5 Goals in Explanations
•  
   6.6 Explainable Systems: Goals and Purposes
•  
   6.7 Operationalizations of Goals
•  
   6.8 How Does This Chapter Inspire Further Directions of XAI?
•  
   6.9 Rapid Access to the Content of This Chapter
•  
   References
•  
  7 Structures Underlying Explanations
•  
   7.1 How Does This Chapter Relate to XAI?
•  
   7.2 The Notion of Structures
•  
   7.3 The Building Blocks: Elementary Process Units and Turns
•  
   7.4 Micro-Level Structures: Elementary Patterns in Turn-Sequences
•  
   7.5 Meso-Level Structures: Pragmatic Frames and ``Jobs'' in Explanation
•  
   7.6 Macro-Level Structures: Dialog Types in Explanation
•  
   7.7 Operationalizations of Structures
•  
   7.8 How Does This Chapter Inspire Further Directions of XAI?
•  
  7.9 Rapid Access to the Content of This Chapter
•  
   7.9.1 The Notion of Structure in Explanations
•  
   7.9.2 The Building Blocks of Explanation: Turns and Elementary Units
•  
   7.9.3 Micro-Level Structures: Elementary Patterns in Turn-Sequences
•  
   7.9.4 Meso-Level Structures: Pragmatic Frames and ``Jobs'' in Explanation
•  
   7.9.5 Macro-Level Structures: Dialog Types in Explanation
•  
   7.9.6 Operationalizing Explanation Structures for XAI
•  
   7.9.7 Future Directions: Towards Socially Intelligent XAI
•  
   References
•  
  8 Roles and Relationships
•  
   8.1 Introduction
•  
   8.2 How Do Roles Relate to sXAI?
•  
   8.3 Roles and Norms
•  
   8.4 Role Relationships
•  
   8.5 Dialog Roles
•  
   8.6 Operationalization
•  
   8.7 Dialog Games
•  
   8.8 Uptake and Roles
•  
   8.9 Rapid Access to the Content of This Chapter
•  
   References
•  
   9 Responsibilities in sXAI
•  
   9.1 How Does This Chapter Relate to XAI?
•  
   9.2 Responsibility and Agency
•  
   9.3 Responsibilities in an Interaction and the Structure of Explaining: A Developmental View
•  
   9.4 Scaffolding Agential Capacities in an Interaction with Marginal Agents
•  
   9.5 Responsibilities for Decisions: A Social Interaction View
•  
   9.6 Operationalizations of Responsibilities and Responsibilities When Designing XAI
•  
   9.7 How Does This Chapter Inspire Further Directions of XAI? And How to Foster Responsibility in the Design of XAI?
•  
   9.8 Rapid Access to the Content of This Chapter
•  
   References
•  
  10 Values and Norms in sXAI
•  
   10.1 How Does This Chapter Relate to XAI?
•  
   10.2 Value-Sensitive Design and Criticisms
•  
   10.3 The Values of sXAI
•  
   10.4 Value-Sensitive Design of Disobedient Robots
•  
   10.5 Conclusion
•  
   10.6 Rapid Access to the Content of This Chapter
•  
   References
•  
   11 Managing the sXAI Interaction: Beliefs, Goals, and Decisional Processes
•  
   11.1 How Does This Chapter Relate to sXAI?
•  
   11.2 Overview of the Chapter
•  
   11.3 The Explanation Process: A Goal-Directed Belief Updating Collaborative Process
•  
   11.4 The Explaining Process: A Multiagent Collaborative Activity
•  
  11.5 sXAI Viewed as Human-Machine Joint Action
•  
   11.5.1 About Human-Human Joint Action
•  
   11.5.2 Cognitive Processes by Which Uncertainty Is Reduced and Coordination Achieved
•  
  11.6 Roles and Task Distribution Between the Human and the Machine
•  
   11.6.1 Human Versus Machine
•  
   11.6.2 About Machines Able to Act as Socially Interactive Agents
•  
   11.7 Goal Management and Associated Decisional Issues in sXAI
•  
   11.8 On Commitment and Its Management
•  
   11.9 Operationalizations of Decisional Processes for Managing the sXAI
•  
   11.10 How Does This Chapter Inspire Further Directions of sXAI?
•  
   11.11 Rapid Access to the Content of This Chapter
•  
   References
•  
   Part II Incrementality
•  
   12 Incremental Communication
•  
   12.1 How Does This Chapter Relate to XAI?
•  
   12.2 Problem Description
•  
  12.3 Human Incremental Communication
•  
   12.3.1 Compositional Account of Incremental HHI
•  
   12.3.2 Standard Account of Incremental HHI
•  
   12.3.3 Processual Account of Incremental HHI
•  
  12.4 Operationalization: Incremental HCI
•  
   12.4.1 Compositional Account of Incremental HCI/sXAI
•  
   12.4.2 Standard Incremental Account of Incremental HCI
•  
   12.4.3 Summary
•  
   12.5 How Does This Chapter Inspire Further Directions of XAI?
•  
   12.6 Rapid Access to the Content of This Chapter
•  
   References
•  
   13 Adaptation
•  
   13.1 How Does This Chapter Relate to XAI?
•  
  13.2 Cognitive Adaptation Processes
•  
   13.2.1 Knowledge Construction Needs Adaptive Interactions
•  
   13.2.2 Examples of Cognitive Adaptation in Explanatory Dialogs
•  
   13.3 Interactive Adaptation
•  
   13.3.1 Adapting the Content
•  
   13.3.2 Explainer Adapting via Scaffolding
•  
   13.3.3 Explainee Adapting via Substantive Contributions
•  
   13.4 Operationalization
•  
   13.5 How Does This Chapter Inspire Further Directions of XAI?
•  
   13.6 Rapid Access to the Content of This Chapter
•  
   References
•  
   14 Models of the Situation, the Explanandum, and the Interaction Partner
•  
   14.1 How Does This Chapter Relate to XAI?
•  
  14.2 Model of the Situation: Context
•  
   14.2.1 Situation Versus Context
•  
   14.2.2 Characteristics and Elements of the Situation
•  
   14.2.3 Computational Representation Approaches
•  
  14.3 Model of the Explanandum
•  
   14.3.1 Mental Representations of the Explanandum
•  
   14.3.2 Construction of a Model of the Explanandum
•  
   14.4 Model of the Interaction Partner
•  
   14.4.1 Dimensions of Partner Models
•  
   14.4.2 Preconditions of Partner Models
•  
   14.4.3 Development of the Partner Model
•  
   14.5 Operationalization
•  
   14.5.1 Knowledge Graphs as a Modeling Tool
•  
   14.5.2 Model of Explanandum
•  
   14.5.3 Model of the Context
•  
   14.5.4 Partner Model
•  
   14.6 How Does This Chapter Inspire Further Directions of XAI?
•  
   14.7 Rapid Access to the Content of this Chapter
•  
   References
•  
   15 Generation of Explanatory Content and Requirements for Social XAI
•  
   15.1 How Does This Chapter Relate to XAI?
•  
   15.2 State-of-the-Art XAI Methods
•  
   15.3 What Questions Do Users Actually Expect to Get Answers on?
•  
   15.4 Operationalization
•  
   15.5 How Does This Chapter Inspire Further Directions of XAI?
•  
   15.6 Rapid Access to the Content of This Chapter
•  
   References
•  
  16 Exploration of Explaining Content
•  
   16.1 How Does This Chapter Relate to XAI?
•  
   16.2 State of the Art of Exploration Strategies
•  
   16.3 Context-Sensitive Exploration
•  
   16.4 Operationalization
•  
   16.5 How Does This Chapter Inspire Further Directions of XAI?
•  
   16.6 Rapid Access to the Content of This Chapter
•  
   References
•  
  17 Interaction History in Social XAI
•  
   17.1 How Does This Chapter Relate to XAI?
•  
   17.2 Taking into Account Repeated Interaction at the Micro-Level
•  
   17.3 Repeated Interaction at the Macro-Level
•  
   17.4 Operationalization
•  
   17.5 How Does This Chapter Inspire Further Directions in XAI?
•  
   17.6 Rapid Access to the Content of This Chapter
•  
   References
•  
   Part III Multimodality
•  
  18 Theoretical Aspects of Multimodal Processing
•  
   18.1 How Does This Chapter Relate to XAI?
•  
   18.2 General Principles and Problems
•  
   18.3 Production and Generation of Speech and Gesture
•  
   18.4 Comprehension of Speech and Gesture
•  
   18.5 Processing of Multimodal Behavior in Social Interactions
•  
   18.6 Operationalization: Multimodality in sXAI
•  
   18.7 Rapid Access to the Content of This Chapter
•  
   References
•  
  19 Characteristics of Nonverbal Behavior
•  
   19.1 How Does This Chapter Relate to XAI?
•  
   19.2 The Dynamics of Nonverbal Behavior as a Challenge for Human–Machine Interaction
•  
   19.3 Forms of Nonverbal Behavior in Human–Human Interaction
•  
   19.3.1 Eye Gaze Behavior
•  
   19.3.2 Co-Speech Gestures
•  
   19.3.3 Adaptors
•  
   19.3.4 Head Movements
•  
   19.3.5 Torso Movements
•  
   19.3.6 Summary: Nonverbal Behavior in Human–Human Interaction
•  
   19.4 Forms of Nonverbal Behavior in Human–Machine Interaction
•  
   19.4.1 Eye Gaze Behavior
•  
   19.4.2 Co-Speech Gestures and Adaptors
•  
   19.4.3 Head Movements
•  
   19.4.4 Torso Movements and Posture
•  
   19.5 Operationalization: Monitoring of Human Nonverbal Behavior in sXAI
•  
   19.6 How Does This Chapter Inspire Further Research on sXAI?
•  
   19.7 Rapid Access
•  
   References
•  
  20 Nonverbal Signals of Affect
•  
   20.1 Relevance of Nonverbal Signals of Affect for XAI
•  
   20.1.1 Monitoring Affective Signals
•  
   20.1.2 Generating Affective Signals
•  
   20.1.3 Inference of Affective Signals
•  
   20.2 Facial Expressions
•  
   20.2.1 Current Use of Facial Expressions in XAI
•  
   20.3 Eyes
•  
   20.3.1 Current Use of Eye Signals in XAI
•  
   20.4 Voice
•  
   20.4.1 Current Use of Voice Signals in XAI
•  
  20.5 Circulatory System
•  
   20.5.1 Heart
•  
   20.5.2 Respiration
•  
   20.5.3 Current Use of Circulatory System Signals in XAI
•  
   20.6 Conclusion
•  
   20.7 Rapid Access to the Content of This Chapter
•  
   References
•  
  21 Ambiguity of Nonverbal Signals
•  
   21.1 How Does This Chapter Relate to XAI?
•  
   21.2 Ambiguity and Nonverbal Behavior
•  
   21.2.1 Factors of Ambiguity in Nonverbal Behavior
•  
   21.2.2 Life Cycle of Ambiguity Resolution
•  
   21.3 Ambiguity in Human-Machine Interactions
•  
   21.3.1 Human-Sender and Agent-Receiver
•  
  21.3.2 Agent-Sender and Human-Receiver
•  
   21.3.2.1 Embodied Agent Explainer
•  
   21.3.2.2 Non-Embodied Agent Explainer
•  
   21.4 Operationalization
•  
   21.5 How Does This Chapter Inspire Further Research on sXAI?
•  
   21.6 Rapid Access
•  
   References
•  
  22 Timing and Synchronization of Multimodal Signals inExplanations
•  
   22.1 How Does This Chapter Relate to XAI?
•  
   22.2 Timing as an Orchestration of Verbal and Nonverbal Events
•  
   22.3 Functions of Multimodal Synchronization
•  
   22.4 Operationalization of Multimodal Synchronization
•  
   22.5 How Does This Chapter Inspire Further Directions of XAI
•  
   22.6 Rapid Access
•  
   References
•  
  23 Multimodality in Explanatory Interactions
•  
   23.1 How Does This Chapter Relate to XAI?
•  
   23.2 Modeling Explanations as Interactions
•  
   23.3 Multimodality in Explanatory Interactions
•  
   23.4 Operationalization of Global Structure and Multimodal Behavior
•  
   23.5 How Does This Chapter Inspire Further Directions of XAI?
•  
   23.6 Rapid Access
•  
   References
•  
   24 Multimodality in Agents
•  
   24.1 How Does This Chapter Relate to XAI?
•  
   24.2 Embodied Agents
•  
   24.2.1 Sense of Others: Explanations in Embodied AI Systems
•  
   24.3 Disembodied Agents
•  
   24.3.1 Embodied Virtual Agents and Multimodality
•  
   24.3.2 Explainable Disembodied Agents
•  
   24.4 Potentials and Risks in Multimodal XAI Implementing Social Cues
•  
   24.4.1 Burdens of Complexity and Its Management
•  
   24.4.2 Potential Bias Amplification and Overtrust in Multimodal Interaction
•  
   24.4.3 Risk of Manipulative Drifts
•  
   24.5 Operationalization
•  
   24.6 How Does This Chapter Inspire Further Research on sXAI?
•  
   24.7 Rapid Access
•  
   References
•  
   25 Visualization and the Use of Multimodality to Explain
•  
   25.1 How Does This Chapter Relate to Social XAI?
•  
   25.2 Visualization in Deep Learning
•  
   25.3 Explainable Deep Learning
•  
  25.4 Visualization in XAI
•  
   25.4.1 Visualization of Post Hoc Explanations
•  
   25.4.2 Other Forms of Visualizations
•  
   25.4.3 Interactivity of Visualizations
•  
   25.5 Multimodality Through Visualization
•  
   25.6 Visual sXAI System—Example
•  
   25.7 Operationalization
•  
   25.8 How Does This Chapter Inspire Further Research on sXAI?
•  
   25.9 Rapid Access
•  
   References
•  
   Part IV Evaluation
•  
  26 Evaluation Principles
•  
   26.1 How Does This Chapter Relate to XAI
•  
   26.2 Measuring the Human Ability as a Social Factor
•  
   26.3 Measuring the Explanation as a Social Factor
•  
   26.4 Measuring the Human Motivation as a Social Factor
•  
   26.5 Measuring the Interaction Sequence as a Social Factor
•  
   26.6 Measuring the Human Spatial Embedding as a Social Factor
•  
   26.7 How Does This Chapter Inspire Further Research in XAI
•  
   26.8 Rapid Access to the Content of This Chapter
•  
   References
•  
  27 Operationalizing Social Interaction
•  
   27.1 How Does This Chapter Relate to XAI?
•  
   27.2 A Basic Operational Model of Social Interactions
•  
   27.3 Refinements and Extensions of the Basic Model
•  
   27.4 Modeling Multimodality
•  
   27.5 Modeling Incrementality
•  
   27.6 Modeling Patterns
•  
   27.7 How Does This Chapter Inspire Further Research?
•  
   27.8 Rapid Access to the Content of This Chapter
•  
   References
•  
  28 Measuring the Outcome of sXAI
•  
   28.1 How Does This Chapter Relate to XAI?
•  
   28.2 Quantitative Measures Applied in XAI
•  
   28.3 Short-Term Subjective Outcomes
•  
   28.4 Short-Term Objective Outcomes
•  
   28.5 Long-Term Outcomes
•  
   28.6 How Does This Chapter Inspire Further Directions of XAI?
•  
   28.7 Rapid Access to the Content of This Chapter
•  
   References
•  
   29 Tasking AI Fairly. How to Empower AI Practitioners With sXAI?
•  
   29.1 How Does This Chapter Relate to XAI?
•  
  29.2 The SyRI Case: Lessons to be Learned from a Political Scandal
•  
   29.2.1 What Went Wrong in the Dutch Fraud Scandal?
•  
   29.2.2 Vulnerability in SyRI
•  
   29.3 From Fair AI to Tasking AI Fairly
•  
   29.4 A Toy Scenario for Ethical Evaluation
•  
   29.4.1 Reflection Questions for an Ethical Assessment
•  
   29.4.2 sXAI to Empower Virtuosic AI Practitioners
•  
   29.5 How Does This Relate to the XAI Community?
•  
   29.5.1 Limits of sXAI
•  
   29.5.2 Future sXAI Development
•  
   29.6 Rapid Access to the Content of This Chapter
•  
   References
•  
  30 The Risk of Manipulation and Deception in sXAI
•  
   30.1 How Does This Chapter Relate to XAI?
•  
   30.2 Philosophical Theories of Manipulation
•  
   30.2.1 Intentional Influence
•  
   30.2.2 The Continuum Model
•  
   30.2.3 The Covertness Model
•  
   30.2.4 Bypassing Rationality
•  
   30.2.5 The Mistake Account
•  
   30.2.6 The Indifference Account
•  
   30.3 Manipulation and Deception in XAI
•  
   30.3.1 The Manipulative Potential of Transparent AI
•  
   30.3.2 Manipulation and Deception by Post Hoc Explanations
•  
  30.4 Does sXAI Aggravate Concerns with Deception and Manipulation?
•  
   30.4.1 Adaptivity as a Medium for Deception and Manipulation
•  
   30.4.2 Social Interaction as a Medium for Deception and Manipulation
•  
   30.5 How Does This Chapter Inspire Further Directions of XAI?
•  
   30.6 Rapid Access
•  
   References