TiiS Special Issue on Human-centered Explainable AI - Call for Papers

[Styliani Kleanthous]

Apologies for cross postings.

Dear Colleagues,

As AI systems increasingly power decision-making in high-stakes domains like healthcare and finance, explainability of AI systems is required for all stakeholders, such as system designers and end users, to make informed decisions. In particular, stakeholders need to know how black-box AI works and how accountable and trustworthy the AI-enabled decisions might be. Although there are existing works on “opening” the black-box of AI through interpretable models, post-hoc explanations, and visualization techniques, there is inadequate work on how to tailor explainable AI to different user groups (inclusivity), how to make explainable AI actionable (empowerment), or how explainable AI helps measure and understand the safety and reliability of AI systems (responsibility). This Special Issue (SI) aims at featuring cutting-edge research in Human-centered Explainable AI (HXAI) around three areas: inclusivity, empowerment, and responsibility.

Topics

This special issue invites submissions that feature original research on the design, development, and evaluation of innovative interactive intelligent systems for explainable AI.  Submissions should directly address one or more of the areas listed below, and demonstrate relevance to TiiS around the two core characteristics of an interactive intelligent system: machine intelligence and user interaction. Interdisciplinary research is highly encouraged.

Specific areas of interest include, but are not limited to:

• Inclusive XAI: design, development, and evaluation of novel interactive explainable AI technologies and systems that address the needs of different user groups, such as data scientists, researchers, end users, administrators, and regulators; individual differences, and culture differences; 
• Empowerment of XAI: design, development, and evaluation of interactive explainable AI technologies and systems that improve user actionability, e.g., knowing when and how to seek recourse or how to incorporate stakeholder voices in design. 
• Responsible XAI: design, development, and evaluation of interactive explainable AI technologies and systems to measure FATE (Fairness, Accountability, Transparency, and Ethics), and demonstrate how such metrics impact the safety and reliability of AI systems.  

Important Dates

• Submission deadline: February 21, 2022
• First-round review decisions: May 15, 2022
• Deadline for revision submissions: June 30, 2022
• Notification of final decisions: September 15, 2022
• Tentative publication: December 2022

Guest Editors:

• UPOL EHSAN, GEORGIA INSTITUTE OF TECHNOLOGY
• STYLIANI KLEANTHOUS, OPEN UNIVERSITY OF CYPRUS
• Q. VERA LIAO, MICROSOFT RESEARCH MONTREAL
• ALISON SMITH-RENNER, DATAMINR
• ADVAIT SARKAR, MICROSOFT RESEARCH CAMBRIDGE
• MARK O. RIEDL, GEORGIA INSTITUTE OF TECHNOLOGY

Editorial Program Committee

• Niels van Berkel, Aalborg University
• Simone Stumpf, University of Glasgow
• Daniel Buschek, University of Bayreuth
• Brian Y Lim, University of Singapore
• Benjamin Tag, University of Melbourne
• Avital Shulner-Tal, University of Haifa
• Philipp Wintersberger, TU Wien
• Tsvi Kuflik, University of Haifa

Submission Information

The journal welcomes articles on any of the above topics in the context of Human-centered Explainable AI. ACM TiiS will encourage original submissions that have not been published or submitted in any form elsewhere, and submissions which may significantly contribute to opening up new and potentially important areas of research and development. ACM TiiS will publish outstanding papers that are "major value-added extensions" of papers previously published in conferences. Such extensions must contribute at least 30% new original work. In this case, authors will need to identify in a separate document the list of extensions over their previously published paper. For more information about submission format and procedure, please refer to the author guidelines.

For questions and further information, please contact the guest editors at tiis-guest-...@acm.org.