Summary of Deepricci: Self-supervised Graph Structure-feature Co-refinement For Alleviating Over-squashing, by Li Sun et al.
DeepRicci: Self-supervised Graph Structure-Feature Co-Refinement for Alleviating Over-squashing
by Li Sun, Zhenhao Huang, Hua Wu, Junda Ye, Hao Peng, Zhengtao Yu, Philip S. Yu
First submitted to arxiv on: 23 Jan 2024
Categories
- Main: Machine Learning (cs.LG)
- Secondary: None
GrooveSquid.com Paper Summaries
GrooveSquid.com’s goal is to make artificial intelligence research accessible by summarizing AI papers in simpler terms. Each summary below covers the same AI paper, written at different levels of difficulty. The medium difficulty and low difficulty versions are original summaries written by GrooveSquid.com, while the high difficulty version is the paper’s original abstract. Feel free to learn from the version that suits you best!
| Summary difficulty | Written by | Summary |
|---|---|---|
| High | Paper authors | High Difficulty Summary Read the original abstract here |
| Medium | GrooveSquid.com (original content) | Medium Difficulty Summary The proposed self-supervised graph structure-feature co-refinement method, DeepRicci, addresses the issue of over-squashing in Graph Neural Networks (GNNs) by leveraging Ricci curvature from Riemannian geometry. This approach introduces a latent Riemannian space to model heterogeneous curvatures and utilizes gyrovector feature mapping for GNNs. The method refines node features through geometric contrastive learning and simultaneously updates graph structure via backward Ricci flow based on a novel differentiable Ricci curvature formulation. Empirical evaluations on public datasets demonstrate the superiority of DeepRicci. |
| Low | GrooveSquid.com (original content) | Low Difficulty Summary DeepRicci is a new way to make Graph Neural Networks work better by using ideas from Riemannian geometry. It’s like solving a puzzle, and this paper shows how it can be done. The method uses something called Ricci curvature to help GNNs learn more effectively. It also helps the network understand different types of graphs better. This approach is tested on real-world data sets and shown to be better than other methods. |
Keywords
* Artificial intelligence * Self supervised
