Call for papers
The community working on dependable AI systems consists of researchers from AI and from software engineering.
This workshop is meant as a place to meet and discuss ongoing and relevant work. We invite submissions from academia as well as industry that raise important questions or propose solutions.
Topics
Topics of interest include:
-
- System architectures of dependable AI systems
- Domain-specific formal languages and logics for modelling functional safety
- Verification of adaptive systems, including runtime monitoring and verification
- Effect of uncertainty
- Accountability, responsibility and liability of AI systems
- Self-explanation, self-criticism and the transparency problem
- Human-machine interaction safety
- Decision quality assurance
- Safety standards and certification
- Evaluation platforms and Standardization
- AI safety education and awareness
- Reports from the field and challenges in dependable AI, including embedded systems/automotive, robotics, health, and industrial processes
Submissions
We invite submission of
- last-minute statements (unpublished, work-in-progress presentations)
- technical papers (full paper/short paper)
- summary papers (short paper)
- application or demo papers.
Technical papers present novel ideas regarding representation and reasoning, and position this result in the context of dependable AI systems. For research-in-progress and late-breaking brilliant ideas, submission of short papers are welcome. We particularly encourage young researchers to submit.
We also invite authors who recently published results of interest to dependable AI systems to submit a short summary paper which highlights the results and explains relevance.
Application contributions present the use and evaluation of software implementing methods that achieve dependability of an AI system, or which present important use cases.
Demos focus on (novel) implementations dependadble AI approaches in software. They can be ‘proof of concepts’, but also mature products. Contributions typically address the ‘extras’ needed to realize the implementation.
All contributions must be submitted via the the instructions on this workshop website. All submissions must be in PDF format and not exceed 6 pages for full papers (plus one for references), and 3 pages for short papers (plus one for references). Formatting instructions and additional information is available on the KI2020 website (LNCS style).
All submissions will be evaluated in a peer review process and selected according to their quality, significance, originality, and potential to generate discussion. Each contribution will be reviewed by at least two referees from the Dependable AI Program Committee. Papers may be accepted for either oral or poster presentation.
The accepted papers will be published as a collection of working papers. As this is a workshop, not a conference, submission of the same paper to conferences (e.g. KI) or journals is acceptable. To accommodate the publishing traditions of different fields, authors of accepted papers can ask that only a one-page abstract of the paper appear in the proceedings. Papers should be formatted according to the LNCS guidelines, available from the KI2020 wepbage, and must be in PDF format. The workshop is also open to people who would like to attend without submitting a paper.