ScaDyG: A New Paradigm for Large-scale Dynamic Graph Learning
Xiang Wu, Xunkai Li, Rong-Hua Li, Kangfei Zhao, Guoren Wang · 2025
Read ScaDyG: A New Paradigm for Large-scale Dynamic Graph Learning by Xiang Wu, Xunkai Li, Rong-Hua Li, Kangfei Zhao, Guoren Wang (2025) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2501-16002.txt.
Description
Wu and co-authors present ScaDyG, a scalable way to learn from graphs that change over time. Dynamic graphs record ties that evolve, with wide real-world use. Most graph networks pass messages and add time tricks, but growing history makes them hard to scale.
ScaDyG first splits past interactions into time steps, within and between steps. That split allows weight-free, time-aware spreading of features during preprocessing. Fine detail inside steps comes from temporal encoding built on exponential functions. A hypernetwork then fuses each node's history adaptively. The design targets industry-scale graphs where older methods bog down in historical data.
Catalog record
| CREATOR | Xiang Wu, Xunkai Li, Rong-Hua Li, Kangfei Zhao, Guoren Wang |
|---|---|
| YEAR | 2025 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 68 KB |
| VERSION | 1.0 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2501-16002