Help me to understand explainable AI ": ExUI – an Explanation Interface to evaluate user-centered XAI
Fuente:
Zenodo
Gespeichert in:
| Hauptverfasser: | , , |
|---|---|
| Format: | Recurso digital |
| Veröffentlicht: |
Zenodo
2025
|
| Schlagworte: | |
| Online-Zugang: | |
| Tags: |
Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
|
| _version_ | 1866902244288364544 |
|---|---|
| author | Schwerk, Anne Emami, Hady Blum, Lothar |
| author_facet | Schwerk, Anne Emami, Hady Blum, Lothar |
| contents | <p>Current explainable AI (XAI) tools lack the essential interactive and adaptive features needed to support a diverse range of users in understanding and applying AI insights effectively. This presentation introduces the Explanation User Interface (ExUI)—an innovative framework designed to address the interpretability challenges of AI-driven systems for sepsis prediction. The ExUI integrates state-of-the-art XAI methodologies, such as counterfactual explanations, feature importance analysis, and temporal visualizations, with dynamic user interfaces tailored for interactive and adaptive experiences. These features ensure that the explanations provided by AI systems are not only technically sound but also actionable, accessible, and aligned with user-specific requirements. This approach is particularly crucial for high-stakes domains, like healthcare, where transparency, ethical use, and decision-making speed are critical. Importantly, the ExUI framework provides a modular and scalable design, making it applicable across a broad range of industries that utilize AI. It directly tackles key challenges in XAI, such as the disagreement problem (variability in explanations across models) and the absence of standardized evaluation metrics. By leveraging adaptive feedback, sophisticated evaluation criteria, and user-centered design principles, ExUI fosters trust, usability, and broader adoption of XAI systems, setting new standards for transparency and accountability in AI applications.</p> |
| format | Recurso digital |
| id | zenodo_https___doi_org_10_5281_zenodo_15170432 |
| institution | Zenodo |
| language | |
| publishDate | 2025 |
| publisher | Zenodo |
| record_format | zenodo |
| spellingShingle | Help me to understand explainable AI ": ExUI – an Explanation Interface to evaluate user-centered XAI Schwerk, Anne Emami, Hady Blum, Lothar XAI User Interface Explanation Interface User-centered AI <p>Current explainable AI (XAI) tools lack the essential interactive and adaptive features needed to support a diverse range of users in understanding and applying AI insights effectively. This presentation introduces the Explanation User Interface (ExUI)—an innovative framework designed to address the interpretability challenges of AI-driven systems for sepsis prediction. The ExUI integrates state-of-the-art XAI methodologies, such as counterfactual explanations, feature importance analysis, and temporal visualizations, with dynamic user interfaces tailored for interactive and adaptive experiences. These features ensure that the explanations provided by AI systems are not only technically sound but also actionable, accessible, and aligned with user-specific requirements. This approach is particularly crucial for high-stakes domains, like healthcare, where transparency, ethical use, and decision-making speed are critical. Importantly, the ExUI framework provides a modular and scalable design, making it applicable across a broad range of industries that utilize AI. It directly tackles key challenges in XAI, such as the disagreement problem (variability in explanations across models) and the absence of standardized evaluation metrics. By leveraging adaptive feedback, sophisticated evaluation criteria, and user-centered design principles, ExUI fosters trust, usability, and broader adoption of XAI systems, setting new standards for transparency and accountability in AI applications.</p> |
| title | Help me to understand explainable AI ": ExUI – an Explanation Interface to evaluate user-centered XAI |
| topic | XAI User Interface Explanation Interface User-centered AI |
| url | https://doi.org/10.5281/zenodo.15170432 |