Special Session Aims and Scope

The fusion of scalable computing infrastructure, big data, and artificial intelligence has boosted the development and application of data science and advanced data analytics. However, the recently emerging threats on the privacy, security, and trust (PST) of the data and the analytics models have shown a dramatically increasing trend with the wide deployment of data analytics applications. Specifically, the PST attacks on data or models such as model inversion attacks, membership inference attacks, data poisoning attacks, evasion attacks, and model backdoors, have severely made advanced data analytics highly vulnerable, particularly in common scenarios where data are distributed or computation is outsourced like MLaaS (Machine Learning as a Service). On the other hand, defence solutions are proposed as new computing schemes, PST frameworks, algorithms, and methods. For example, differential privacy, federated learning, and machine unlearning are proposed for privacy protection in data analytics, and adversarial machine learning is proposed to achieve robust, secure, and trustworthy data analytics. Given the importance and urgency, this special issue aims to provide a venue for researchers, practitioners and developers from different background areas relevant to PST and data analytics to exchange their latest experience, research ideas, and synergic research and development on fundamental issues and applications about privacy, security, and trust issues in data analytics, as a strong supplement to the main track of data science and advanced analytics.

This special session mainly focuses on the discussions of privacy, security, and trust in data analytics, which generally covers (but not limited to) the topics in privacy-preserving technology, privacy attacks, federated learning, machine unlearning, data poisoning attacks, model evasion attacks, adversarial learning, model robustness, secure machine learning integrating cryptographic techniques, blockchain techniques protection PST of data and models, etc.

Accepted research papers will be included in the ADMA 2025 proceedings.

Topics of Interest

This special session invites authors to submit original manuscripts that demonstrate and explore current advances in all related areas mentioned above. Topics of interest include, but are not limited to:

  • New privacy, security and trust opportunities and challenges in data analytics
  • Novel theories and modelling for privacy, security, and trust in data analytics
  • Private, secure, and trust deep learning for data analytics
  • Privacy-preserving data mining and machine learning
  • Federated/collaborative learning
  • Machine unlearning
  • Adversarial machine learning for robust data analytics
  • Transfer learning for private, secure, and trust data analytics
  • Data poisoning and model evasion attacks and defences
  • Cryptographic techniques based private, secure, and trust data analytics
  • Privacy, security, and trust management for data analytics
  • Emerging approaches and strategies for the security analysis of IoT devices
  • Blockchain for privacy, security, and trust in data analytics
  • Real-world applications for private, secure and trust data analytics
  • Generative AI, including LLMs, for private, secure and trust data analytics
  • Privacy, security and privacy issues, trends, and challenges in data analytics

Submission Guideline and Reviewing

CMT Acknowledgment:

The Microsoft CMT service was used for managing the peer-reviewing process for this conference. This service was provided for free by Microsoft and they bore all expenses, including costs for Azure cloud services as well as for software development and support.

Submission Site: https://cmt3.research.microsoft.com/ADMA2025

Formatting Guidelines:

  1. The paper should be in English and contain unpublished contributions to the data mining and related fields.
  2. Manuscripts must be prepared in accordance with the LNAI (Lecture Notes in Artificial Intelligence) format. For the template and details on the LNCS style, see Springer's Author Instructions.
  3. The paper should NOT exceed 15 pages in LNAI format.
  4. Submissions are reviewed in a double-blind manner. For ADMA2025, the double-blind means that:
    a. Author identities and affiliations are not disclosed to reviewers during the review process.
    b. Authors should prepare and submit suitably blinded manuscripts that do not reveal author and affiliation information. Specific requirements to this end are detailed below.
    c. Authors and reviewers alike make an honest effort to avoid accidentally de-blinding any submission.
  5. The list of authors at the time of submission is final and cannot be changed.
  6. "Submitted papers" must comply with all of the rules below. Any violation may result in an "desk reject".

Manuscript Preparation:

  1. Authors must submit PDFs without author names and affiliations. Submitted PDFs must also not contain any metadata that could reveal author identities or affiliations.
  2. Authors should aim to avoid copy&pasting substantial amounts of text from their own prior publications, as such text blocks may be readily recognized by experts familiar with the state of the art and recent papers in the area.
  3. Similarly, authors should not directly reuse figures from their own prior publications without attribution. Ideally, “fresh” figures should be prepared and used whenever possible. If that is not a viable option, that is, if a figure must be reused, then a citation should be included giving credit to the original paper from which the illustration has been adapted.
  4. Authors must refrain from using any specific formatting tricks, linguistic mannerisms, figure styles, or other stylistic idioms that could hint at or disclose the author identity or affiliation.
  5. Submitted papers should not include an acknowledgements section or funding acknowledgements (even when blinded) since it can indicate the country of residence of (some of) the authors. (Such acknowledgements of course may be added to the camera-ready version.)
  6. Authors who seek to refer to an online appendix or to supplemental materials (e.g., source code, videos, etc.) may still do so. However, instead of directly providing a URL or tech report number, authors should include a note that the appendices and/or supplemental materials in question are available from the track chair upon request.
    a. Authors must provide all supplemental materials and/or appendices that a submission refers to in blinded form and sent them to the PC chair by email (or other approaches appointed by the PC chair) before the submission deadline.
    b. The track chair will retain a copy of all submitted materials for the duration of the reviewing process. Any such provided appendices or supplemental materials are not subject to review and may be consulted by the reviewers at their own discretion.
  7. It is imperative to acknowledge the contributions of AI models in the generation of textual content. Authors utilizing AI-generated text in their manuscripts are required to assume full responsibility for the accuracy, integrity, and originality of the material presented. Furthermore, any section of the paper employing AI-generated text should include clear documentation and description of the AI system utilized. This transparency ensures that readers can discern between human-authored content and text generated by AI, fostering a culture of accountability and integrity within the scholarly community.

Own Prior Work, Well-Known Projects, and Research Artifacts

  1. Authors should not upload their manuscript to preprint servers (such as arXiv) or their personal websites while the paper is under review, or otherwise publicly reveal their authorship of the manuscript under review.
  2. As an exception to the previous rule, if a prior version of the manuscript has already been uploaded to a public preprint server prior to submission (e.g., if the paper is a re-submission of a paper previously rejected at another single-blind conference), then the paper may still be submitted to ADMA2025. However, such papers must be blinded when submitted to ADMA2025.
  3. In exceptional circumstances that force a violation of the above two rules (e.g., a technical report or thesis must be filed in order for a student to graduate), the authors should contact the Program Chair prior to publicizing their manuscript content to avoid misunderstandings.
  4. When submitting or extending a prior workshop publication, the workshop paper is treated as an online preprint for the purpose of the double-blind peer-review process. However, authors must proactively disclose the existence of a prior workshop version of a submitted paper. Such information should be emailed to the Program Chair. Failure to disclose a prior workshop publication is considered self-plagiarism.
  5. After communicating with the Program Chair, submissions with conditions 2., 3., or 4. above are recommended to add a headline (or footer) on the first page with the following information: “This submission is based on Preprints, Paper Announcements, and Prior Workshop Papers. The information has been communicated with the Program Chair. The author(s) and the Program Chair request the reviewers not to actively search for the author names to ensure fairness of the double-blind review process.” Please contact the Program Chair at least 24 hours before the submission deadline if this applies to your submission. If the paper is accepted, the above text should be removed and appropriate citations if any should be added (e.g., citation for prior workshop paper, technical report or thesis and its relation to the accepted paper).
  6. If your submission is an extended version of a workshop paper with DOI, you will be asked to provide a blinded version of the workshop paper as supplementary material. This will be used by the reviewers to verify that there is sufficient amount of new material in the extended version to warrant a publication at ADMA2025.