Graph Neural Networks Applications Across Domains: All Insights You Need
Abderaouf Bahi iD in · 2026
Read Graph Neural Networks Applications Across Domains: All Insights You Need by Abderaouf Bahi iD in (2026) 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-2606-27202.txt.
Introduction
Bahi surveys graph neural networks, models that learn from data wired together as networks of nodes and links. Spectral and spatial formulations are derived from common first principles. Expressive power is tied to the Weisfeiler-Leman hierarchy, through precise statements of what today's architectures can and cannot tell apart. Twelve application domains are then examined in this review from 2026, from drug discovery to climate modelling.
Each domain gets its graph-building choices, costs, and dominant architecture families. Reported gains are separated from artefacts of weak baselines or lucky data splits. Cross-domain patterns recur: mixed-neighbour graphs and sheer scale undermine the same models almost everywhere. Time-changing graphs stay harder than static ones, and leaderboard winners rarely reach deployment. Over-smoothing, robustness, fairness, and explainability are treated as adoption constraints throughout. Claims for graph foundation models are judged suggestive, not settled.
Catalog record
| CREATOR | Abderaouf Bahi iD in |
|---|---|
| YEAR | 2026 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 410 KB |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2606.27202