Semantic Interoperability of Sensitive Data in Federated AI (SISFAI)

In conjunction with EDOC 2021 Enterprise Computing Conference

Workshop date: 25 October 2021

Join here for online meeting with (MS Teams)

Background and goals

In the Industry 4.0 era, digital enterprises generate and use vast amounts of data that are captured and stored at different locations to improve their business models by leveraging Artificial Intelligence (AI). AI systems that rely on central data processing are not suitable in most applications since central data processing leads to issues on data accessibility, standardization, privacy, and trustworthiness. Furthermore, semantic interoperability plays a fundamental role in applications where data have become increasingly diverse and complex. AI systems may have biased or even potentially dangerous results if the meaning of data is not well understood.

This workshop aims to address the semantic interoperability of federated data integration approaches for advanced analytics provided by AI applications that rely on different conditions for data accessibility. These approaches may include AI concerns on privacy, trust, resilience, digital ethics, law, and human rights and should follow the FAIR principles to foster not only higher levels of semantic interoperability but also findability, accessibility, and reusability.

Topics of Interest

The topics of interest include, but are not limited to, the following:

Workshop program on October 25th (Monday), CET:

09:30h - 09:45h: Opening
09:45h - 10:30h: Keynote: "The FAIR principles and semantic interoperability" (Luiz Olavo Bonino, University of Twente)
10:30h - 11:00h: Break
11:00h - 12:00h: Paper presentations

  • Paper: A diagnostic of information security features of Eletronic Data Capture Systems in the Personal Health Train context. Zihan Wang, Wallace Ugulino and João Moreira
  • Paper: Towards Semantic Description of Explainable Machine Learning Workflows. Patricia Inoue Nakagawa, Luís Ferreira Pires, João Moreira and Luiz Olavo Bonino

    • 12:00h - 12:15h: Break
    12:15h - 13:00h: Paper presentation and open discussion
  • Paper: Sec4ML: An approach to support Cybersecurity Data Publishing for Machine Learning tasks. Madalena Silva, Kelli Cordeiro and Maria Claudia Cavalcanti

    • Submission and publication

    Papers can be submitted via the workshop website hosted on EasyChair (https://easychair.org/conferences/?conf=sisfai2021). Submitted papers should be in PDF format and follow the IEEE Proceedings Formatting Guideline (6-10 pages). The papers will be evaluated based on their scientific and technical contribution as well as their appropriateness, significance, and clarity by at least two members of the Program Committee. Each accepted paper that will be presented at the workshop must have one of its authors registered to EDOC’21 before the camera-ready deadline.

    Accepted and presented papers will be published by the IEEE Computer Society Press in a workshop proceedings and made available through IEEE Xplore and the IEEE Digital Library. Selected papers will be invited to submit an extended version to the special issue on Privacy, Trust and Fairness in Data of MDPI's Applied Science journal.

    Authors should submit the camera-ready version for both EasyChair and the IEEE CPS service: link for the camera ready submission is available under "Important Dates" on the conference web page.

    Concerning the presentation of your paper, it is recommended that it should be presented in the format of a 25-30 min. oral presentation (including discussion). Please prepare it according to the instructions available at the EDOC conference website.

    Workshop Format: please note that all events at the EDOC 2021 will take place online due to the covid-19 pandemic and the travel restrictions related to it.

    Important dates

    *The camera ready version and registration for papers accepted in the two-stage submission process to the TEAR workshop at CBI are also due 27 Sept.

    Workshop Organizers

    Program Committee