A Diffusive Data Augmentation Framework for Reconstruction of Complex Network Evolutionary History

En Xu, Can Rong, Jingtao Ding, Yong Li

Keywords: Complex Network, Temporal Network, Diffusion Model, Data Augmentation

TL;DR: A Diffusive Data Augmentation Framework for Reconstruction of Complex Network Evolutionary History

Abstract: The evolutionary processes of complex systems contain critical information about their functional characteristics. The generation time of edges can reveal the historical evolution of various networked complex systems, such as protein-protein interaction networks, ecosystems, and social networks. Recovering these evolutionary processes holds significant scientific value, such as aiding in the interpretation of the evolution of protein-protein interaction networks. However, the scarcity of temporally labeled network data poses challenges for predicting edge generation times under current network structures, leading to issues of insufficient data and significant differences between training and prediction networks. To address this, we introduce a diffusion model that learns the generative mechanisms of networks, producing sufficient augmented network data to effectively mitigate issues of limited and incomplete data. Experimental results demonstrate a 13.7% improvement in prediction accuracy using our approach. Moreover, the model can uniformly predict edge generation times across different types of networks, eliminating the need to retrain the model for each specific network, thus significantly enhancing generalization capability and efficiency.

Primary Area: applications to physical sciences (physics, chemistry, biology, etc.)

Code Of Ethics: I acknowledge that I and all co-authors of this work have read and commit to adhering to the ICLR Code of Ethics.

Submission Guidelines: I certify that this submission complies with the submission instructions as described on https://iclr.cc/Conferences/2025/AuthorGuide.

Reciprocal Reviewing: I understand the reciprocal reviewing requirement as described on https://iclr.cc/Conferences/2025/CallForPapers. If none of the authors are registered as a reviewer, it may result in a desk rejection at the discretion of the program chairs. To request an exception, please complete this form at https://forms.gle/Huojr6VjkFxiQsUp6.

Anonymous Url: I certify that there is no URL (e.g., github page) that could be used to find authors’ identity.

No Acknowledgement Section: I certify that there is no acknowledgement section in this submission for double blind review.

Submission Number: 5462