About

Structured data including graphs, text, and time series is foundational to web technologies and applications, plays a critical role in fields like network science, recommendation systems, natural language processing, and multimedia. Recently, Optimal Transport (OT) has gained considerable traction in machine learning as a powerful framework for measuring the similarity between complex probability distributions, providing valuable tools for structured web data applications, especially in graph and network analysis and transfer learning on web data. In addition to the success of OT on modeling structured data, recent development in generative models, foundation models, and large language models brings new research opportunities of OT for structured data generation, making it a more vibrant area for research and innovation.

This workshop aims to provide a premier interdisciplinary forum for researchers from different communities to discuss the latest trends, innovations, applications, and challenges of OT and structured data modeling and generation, with a focus on web-related applications. By highlighting OT's role in recommendation systems, network analysis, graph learning, and other web technologies, the workshop of Optimal Transport for Structured Data Modeling and Generation (OT-SDMG) is to foster collaboration and drive forward the development of OT methodologies that can tackle the complexities of structured data in web contexts.

Call for Papers

We welcome contributions of both technical and perspective papers from a wide range of topics, including but not limited to the following topics of interest:

1. Theoretical Foundations and Computational Advances in OT:

   • Recent developments in OT theory (e.g., Sliced Wasserstein distances, Gromov-Wasserstein distances, and Unbalanced and partial OT).

   • Scalable OT algorithms for large-scale structured data.

   • Multi-GPU and hardware-accelerated implementations of OT.

2. OT for Structured Data Modeling:

   • OT-based applications such as graph analysis, natural language processing, computer vision, bioinformatics.

   • OT-based methods for multi-modal structured data, e.g, text, images, videos, and audios.

   • OT-based domain adaptation and transfer learning approaches.

3. OT for Structured Data Generation:

   • Integration of OT into generative models (e.g., GANs, VAEs, diffusion models) for web applications.

   • Leveraging OT in foundation models (e.g., LLMs, LVLMs) for structured data modeling, generation, and classification across multiple domains and modalities.

   • Applications of OT in web technologies, including recommendation systems, semantic search, and automated content generation.

4. Other Web Applications of OT:

   • E.g., recommendation systems, semantic search, and automated content generation.

Paper Submission Guidelines:

We invite the submission of papers with 4-6 pages. References will not count towards the page limit. Workshop papers must be written in English, in double-column format, and must adhere to the ACM template and formatting (also available in Overleaf).

Submissions will be peer-reviewed, single-blinded, and assessed based on their novelty, technical quality, significance, clarity, and relevance regarding the workshop topics. Submissions introducing interesting experimental phenomena and open problems of optimal transport and structured data modeling are welcome as well. All accepted papers will be posted on the workshop website.

Submission is permitted for papers that are under review, posted in arXiv, or planned to be submitted elsewhere. To encourage original works, submissions that are already accepted for another conference or a journal are not accepted. This policy also applies to submissions that overlap substantially in technical content with papers previously published or accepted.

Authors of accepted papers must confirm whether they agree/want to have their papers to be included in the Companion Proceedings of WWW 2025. Papers included in the Companion Proceedings may not be submitted to other conferences/journals (subject to their policies). Authors can confirm their choice in the submission page of OpenReview. The accepted papers (except for papers submitted from the Workshop Fast-Track, please see below) will not be included in the Companion Proceedings of WWW 2025 so that they can be resubmitted to other conferences/journals.

The submission website is OTSDMG 2025.

Workshop Fast-Track Paper Submission Guidelines:

Our workshop participates in the Workshop Fast-Track, which is for papers rejected by the main conference of TheWebConf2025 to be resubmitted to the workshops. The guidelines for converting a main conference submission to a workshop one are as follows.

The authors need to add below References an Appendix in which they will cut and paste reviews from the main review process. In Appendix authors should also include one section called “Improvements” in which they briefly describe if they were able to revise some of the issues voiced in the main review, and how. Only light revisions are expected and no revisions where the work is in good shape are OK too. Heavy changes are not allowed. Also, tweaking review comments is not allowed.

The reviews from the main process will be used by the workshop organizers to decide whether the paper will be accepted and will not be shared with any third parties.

By default accepted papers from the fast-track will be published in the ACM Companion proceedings. If the authors wish to opt out their work from the proceedings (e.g., they want to resubmit it to other venues), they can inform the workshop organizers upon acceptance.

Workshop Fast-Track Timeline: Submission period Jan 20-26th 11.59 AOE, Decision notification Jan 27th AOE.

Schedule

TBD

Invited Speaker

Important Days

Submission Deadline: January 1, 2025 February 15, 2025 (Anywhere on Earth)
Paper notification: March 1, 2025
Camera-ready Deadline: March 15, 2025
Workshop Fast-Track Timeline: Please see paper submission guidelines
Workshop Day: April 29, 2025