9-13 September 2024

1st Workshop on 

Mining and Learning Real-world Dynamics via High-order Networks

ABSTRACT

Traditionally, complex systems have been successfully studied through graphs abstracting the underlying relations with vertices and edges connecting pairs of interacting components. Yet, many real-world systems are characterized by group interactions that cannot be described simply in terms of dyads. Studying such systems requires new mathematical frameworks and scientific methodologies for its investigation. Hypergraphs are the perfect candidates to tackle this task, as these structures are a generalization of graphs where a (hyper)edge allows the connection of an arbitrary number of vertices. 

Given the considerable potential and the growing interest of the research community in overcoming the challenges of exploiting hypergraphs and, more generally, high-order networks to unravel group relations, we propose a workshop on Mining and Learning Real-world Dynamics via High-order Networks (MLH) at ECML-PKDD 2024. This workshop is dedicated to exploring real-world dynamics through the lens of high-order networks, specializing in the analysis of complex relationships. Its primary objective is to cultivate an environment conducive to advancing learning techniques tailored for modeling higher-order relationships, such as group dynamics and interactions. These relationships are essential for comprehending a wide range of phenomena in social, natural, and economic systems. 

The goal of this workshop is to focus the attention of the ECML-PKDD research community on addressing the open questions and challenges in this thriving research area. Given the broad range of competencies in the ECML-PKDD community, the workshop will welcome foundational contributions as well as applications to real-world problems, tool implementations, and the development of benchmark aspects of high-order network learning.

TOPICS AND THEMES

This workshop focuses on the theme of assessing and improving the state-of-the-art on high-order network mining and learning. Our venue welcomes submissions of theoretical and applicative contributions in these areas as well as the proposal of benchmarking tools. The workshop also encourages surveys and vision papers that review particular facets of the hypernetwork mining and learning tasks, critically examine ongoing challenges, and highlight promising research directions. Additionally, we particularly invite multidisciplinary work that blends ideas from data mining, machine learning, and computer science with social and natural sciences, aiming to address current societal challenges.

The topics include (but are not limited to):

SUBMISSION GUIDELINES

We invite submissions of original research on all aspects of high-order network mining and learning (see a non-exhaustive list of topics above). Each accepted paper will be included in the workshop proceedings (published by Springer Communications in Computer and Information Science) and presented in the talk session. Authors will have the faculty to opt-in or opt-out. 

Submitted papers should not exceed 12 pages (excluding references) and must be self-contained, written in English, and formatted according to the Springer LNCS guidelines. 

The review process is double-blind and all papers need to be "best-effort" anonymized. We strongly encourage making code and data available anonymously (e.g., in an anonymous GitHub repository via Anonymous GitHub or in a Dropbox folder). Submissions will be evaluated by at least two reviewers on the basis of relevance, technical quality, potential impact, and clarity.

Additional details will follow.

IMPORTANT DATES

Header image credits: Antanasc