Human-Centered Explainable AI (HCXAI): Beyond Opening the Black-Box of AI

Ehsan, Upol; Wintersberger, Philipp; Liao, Q. Vera; Watkins, Elizabeth Anne; Manger, Carina; Daumé, Hal; Riener, Andreas; Riedl, Mark O.

doi:10.1145/3491101.3503727

Cited by 61 publications

(21 citation statements)

References 15 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…In recent years, the explainable AI (XAI) community has made tremendous progress at developing techniques for explaining how AI systems work [7,30,57,58,101]. Much of the work in XAI has focused on discriminative algorithms: how they generally make decisions (e.g.…”

Section: Design For Explanationsmentioning

confidence: 99%

“…Recent work in human-centered XAI (HCXAI) has emphasized designing explanations that cater to human knowledge and human needs [30]. This work grew out of a general shift toward human-centered data science [6], in which the import of explanations is not for a technical user (data scientist), but for an end user who might be impacted by a machine learning model.…”

Section: Design For Explanationsmentioning

confidence: 99%

“…Finally, to support learning how to effectively interact with a generative AI application, designers ought to help users create accurate mental models [54] through explanations [7,30,57,58,101]. Explanations can help answer general questions such as what a generative AI application is capable or not capable of generating, how the model's controls impact its output, and how the model was trained and the provenance of its training data.…”

Section: Designing For User Aimsmentioning

confidence: 99%

See 2 more Smart Citations

Toward General Design Principles for Generative AI Applications

Weisz¹,

Müller²,

He³

et al. 2023

Preprint

View full text Add to dashboard Cite

Fig. 1. Seven principles for the design of generative AI systems. Six of these principles are presented in overlapping circles, indicating their relationships to each other. One principle stands alone, the directive to design against potential harms that may be caused by a generative model's output, misuse, or other harmful effects. These principles are bounded in an environment of generative variability, in which the outputs of a generative AI application may vary in quantity, quality, character, or other characteristics.Generative AI technologies are growing in power, utility, and use. As generative technologies are being incorporated into mainstream applications, there is a need for guidance on how to design those applications to foster productive and safe use. Based on recent research on human-AI co-creation within the HCI and AI communities, we present a set of seven principles for the design of generative AI applications. These principles are grounded in an environment of generative variability. Six principles are focused on designing for characteristics of generative AI: multiple outcomes & imperfection; exploration & control; and mental models & explanations. In addition, we urge designers to design against potential harms that may be caused by a generative model's hazardous output, misuse, or potential for human displacement. We anticipate these principles to usefully inform design decisions made in the creation of novel human-AI applications, and we invite the community to apply, revise, and extend these principles to their own work.CCS Concepts: • Human-centered computing → HCI design and evaluation methods; Interaction paradigms; HCI theory, concepts and models.

show abstract

Section: Design For Explanationsmentioning

confidence: 99%

Section: Design For Explanationsmentioning

confidence: 99%

Section: Designing For User Aimsmentioning

confidence: 99%

See 1 more Smart Citation

Toward General Design Principles for Generative AI Applications

Weisz¹,

Müller²,

He³

et al. 2023

Preprint

View full text Add to dashboard Cite

show abstract

“…However, due to the inherent human-centric property of explainability (i.e., explanations are only successful if they match the specific needs of the person receiving them), there is no one-size-fits-all solution in the growing collection of XAI techniques (Liao and Varshney 2021). Consequently, more and more researchers start to work with human-centered explainable artificial intelligence (HCXAI) (Ehsan and Riedl 2020;Wang et al 2019;Liao and Varshney 2021;Ehsan et al 2022), putting the human at the center of technology design (Ehsan and Riedl 2020). AI systems have become ubiquitous in intelligent applications around our daily life, and involve nearly everyone as stakeholder rather than experts only.…”

Section: Human-centered Explainable Aimentioning

confidence: 99%

It Is like Finding a Polar Bear in the Savannah! Concept-Level AI Explanations with Analogical Inference from Commonsense Knowledge

He¹,

Balayn²,

Buijsman³

et al. 2022

HCOMP

View full text Add to dashboard Cite

With recent advances in explainable artificial intelligence (XAI), researchers have started to pay attention to concept-level explanations, which explain model predictions with a high level of abstraction. However, such explanations may be difficult to digest for laypeople due to the potential knowledge gap and the concomitant cognitive load. Inspired by recent work, we argue that analogy-based explanations composed of commonsense knowledge may be a potential solution to tackle this issue. In this paper, we propose analogical inference as a bridge to help end-users leverage their commonsense knowledge to better understand the concept-level explanations. Specifically, we design an effective analogy-based explanation generation method and collect 600 analogy-based explanations from 100 crowd workers. Furthermore, we propose a set of structured dimensions for the qualitative assessment of analogy-based explanations and conduct an empirical evaluation of the generated analogies with experts. Our findings reveal significant positive correlations between the qualitative dimensions of analogies and the perceived helpfulness of analogy-based explanations. These insights can inform the design of future methods for the generation of effective analogy-based explanations. We also find that the understanding of commonsense explanations varies with the experience of the recipient user, which points out the need for further work on personalization when leveraging commonsense explanations.

show abstract

“…Machine learning developers have created a large number of explanation techniques for various types of models [4,12,2,45], and the effects of XAI on user understanding has been subject of several user studies in the AI literature [3,47,55,8,11]. However, despite efforts to create benchmarks for objectively evaluating XAI techniques [16,30,56,1], understanding how exactly XAI affects trust and behavior of lay users in human-AI interaction has remained a challenge [44,12,20,21].…”

Section: Related Workmentioning

confidence: 99%

Reassuring, Misleading, Debunking: Comparing Effects of XAI Methods on Human Decisions

Humer¹,

Hinterreiter²,

Leichtmann³

et al. 2022

Preprint

View full text Add to dashboard Cite

Trust calibration is essential in AI-assisted decision-making tasks. If human users understand the reasons for a prediction of an AI model, they can assess whether or not the prediction is reasonable. Especially for high-risk tasks like mushroom hunting (where a wrong decision may be fatal), it is important that users trust or overrule the AI in the right situations. Various explainable AI methods are currently being discussed as potentially useful for facilitating understanding and to calibrate user trust. So far, however, it is unclear which approaches are most effective. Our work takes on this issue; in a between-subjects experiment with 𝑁 = 501 participants. Participants were tasked to classify the edibility of mushrooms depicted on images. We compare the effects of three XAI methods on human AI-assisted decision-making behavior: (i) Grad-CAM attributions; (ii) nearest neighbor examples; and (iii) an adoption of network dissection. For nearest neighbor examples, we found a statistically significant improvement in user performance compared to a condition without explanations. Effects did not reach statistical significance for Grad-CAM and network dissection. For the latter, however, the effect size estimators show a similar tendency as for nearest neighbor. We found that the effects also varied for different task items (i.e., mushroom images). Explanations seem to be particularly effective if they reveal possible flaws in case of wrong AI classifications or reassure users in case of correct classifications. Our results suggest that well-established methods might not be as beneficial to end users as expected and that XAI techniques must be chosen carefully in real-world scenarios.

show abstract

Human-Centered Explainable AI (HCXAI): Beyond Opening the Black-Box of AI

Cited by 61 publications

References 15 publications

Toward General Design Principles for Generative AI Applications

Toward General Design Principles for Generative AI Applications

It Is like Finding a Polar Bear in the Savannah! Concept-Level AI Explanations with Analogical Inference from Commonsense Knowledge

Reassuring, Misleading, Debunking: Comparing Effects of XAI Methods on Human Decisions

Contact Info

Product

Resources

About