Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

Explaining AI Decisions: Towards Achieving Human-Centered Explainability in Smart Home Environments

Md Shajalal, Alexander Boden, Gunnar Stevens, Delong Du, Dean-Robin Kern

TL;DR

This paper addresses the problem of explainability for AI-driven smart home decisions, arguing that current explanations are often too technical for lay users and can hinder adoption. It surveys technical XAI methods (global/local, model-agnostic/specific) and human-centered approaches, highlighting a gap between advanced explanations and user-centric understanding. Through experiments on energy demand forecasting (REFIT) and personal thermal comfort prediction (UC Berkeley wearable data), it shows that popular XAI techniques like SHAP and DeepLIFT yield explanations that are difficult for non-experts to interpret or act upon. The authors advocate a human-centered XAI approach guided by HCI methods—user studies, prototyping, technology probes analysis, and heuristic evaluation—to design explanations that are understandable, trustworthy, and actionable for smart-home occupants, with implications for better adoption and energy efficiency.

Abstract

Smart home systems are gaining popularity as homeowners strive to enhance their living and working environments while minimizing energy consumption. However, the adoption of artificial intelligence (AI)-enabled decision-making models in smart home systems faces challenges due to the complexity and black-box nature of these systems, leading to concerns about explainability, trust, transparency, accountability, and fairness. The emerging field of explainable artificial intelligence (XAI) addresses these issues by providing explanations for the models' decisions and actions. While state-of-the-art XAI methods are beneficial for AI developers and practitioners, they may not be easily understood by general users, particularly household members. This paper advocates for human-centered XAI methods, emphasizing the importance of delivering readily comprehensible explanations to enhance user satisfaction and drive the adoption of smart home systems. We review state-of-the-art XAI methods and prior studies focusing on human-centered explanations for general users in the context of smart home applications. Through experiments on two smart home application scenarios, we demonstrate that explanations generated by prominent XAI techniques might not be effective in helping users understand and make decisions. We thus argue for the necessity of a human-centric approach in representing explanations in smart home systems and highlight relevant human-computer interaction (HCI) methodologies, including user studies, prototyping, technology probes analysis, and heuristic evaluation, that can be employed to generate and present human-centered explanations to users.

Explaining AI Decisions: Towards Achieving Human-Centered Explainability in Smart Home Environments

TL;DR

This paper addresses the problem of explainability for AI-driven smart home decisions, arguing that current explanations are often too technical for lay users and can hinder adoption. It surveys technical XAI methods (global/local, model-agnostic/specific) and human-centered approaches, highlighting a gap between advanced explanations and user-centric understanding. Through experiments on energy demand forecasting (REFIT) and personal thermal comfort prediction (UC Berkeley wearable data), it shows that popular XAI techniques like SHAP and DeepLIFT yield explanations that are difficult for non-experts to interpret or act upon. The authors advocate a human-centered XAI approach guided by HCI methods—user studies, prototyping, technology probes analysis, and heuristic evaluation—to design explanations that are understandable, trustworthy, and actionable for smart-home occupants, with implications for better adoption and energy efficiency.

Abstract

Smart home systems are gaining popularity as homeowners strive to enhance their living and working environments while minimizing energy consumption. However, the adoption of artificial intelligence (AI)-enabled decision-making models in smart home systems faces challenges due to the complexity and black-box nature of these systems, leading to concerns about explainability, trust, transparency, accountability, and fairness. The emerging field of explainable artificial intelligence (XAI) addresses these issues by providing explanations for the models' decisions and actions. While state-of-the-art XAI methods are beneficial for AI developers and practitioners, they may not be easily understood by general users, particularly household members. This paper advocates for human-centered XAI methods, emphasizing the importance of delivering readily comprehensible explanations to enhance user satisfaction and drive the adoption of smart home systems. We review state-of-the-art XAI methods and prior studies focusing on human-centered explanations for general users in the context of smart home applications. Through experiments on two smart home application scenarios, we demonstrate that explanations generated by prominent XAI techniques might not be effective in helping users understand and make decisions. We thus argue for the necessity of a human-centric approach in representing explanations in smart home systems and highlight relevant human-computer interaction (HCI) methodologies, including user studies, prototyping, technology probes analysis, and heuristic evaluation, that can be employed to generate and present human-centered explanations to users.
Paper Structure (15 sections, 6 figures, 2 tables)

This paper contains 15 sections, 6 figures, 2 tables.

Table of Contents

  1. Introduction
  2. Human-Centered XAI and Current Progress
  3. Technical XAI
  4. Human-centered XAI
  5. The Research Gap
  6. Human-Centered Explainability in Smart Home
  7. Household energy demand forecasting
  8. Occupants' thermal comfort preference modeling
  9. Experiments and Analysis
  10. Energy demand forecasting in smart home
  11. Personal thermal comfort preference prediction
  12. Challenges and HCI Techniques for Human-centered Explainability
  13. Challenges in achieving human-centered explainability
  14. HCI techniques to enhance human-centered explainability
  15. Conclusions and Future Directions

Figures (6)

  • Figure 1: An overview of a human-centered XAI-enabled Smart Home systems
  • Figure 2: Explanations for weekly energy demand forecasting highlighting the contributions of different appliances.
  • Figure 3: Explanations for weekly energy demand forecasting highlighting consumption activity corresponding to the time (day)
  • Figure 4: Global explanation for personal thermal comfort preference prediction highlighting model's overall priorities.
  • Figure 5: Explanation for a decision that predicts the occupants felling "warmer".
  • ...and 1 more figures