Summary of Adafgl: a New Paradigm For Federated Node Classification with Topology Heterogeneity, by Xunkai Li et al.
AdaFGL: A New Paradigm for Federated Node Classification with Topology Heterogeneity
by Xunkai Li, Zhengyu Wu, Wentao Zhang, Henan Sun, Rong-Hua Li, Guoren Wang
First submitted to arxiv on: 22 Jan 2024
Categories
- Main: Machine Learning (cs.LG)
- Secondary: Artificial Intelligence (cs.AI); Databases (cs.DB); Social and Information Networks (cs.SI)
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 Recently, Federated Graph Learning (FGL) has gained attention for breaking data silos using graph neural networks. Current FGL studies simulate semi-supervised node classification by dividing the global graph into client subgraphs. However, real-world implementations often exhibit varying topologies due to local data engineering, leading to heterogeneity challenges in FGL. Unlike label Non-iid problems, FGL heterogeneity arises from topological divergence among clients, revealing homophily or heterophily. To address this, we introduce structure Non-iid split and propose AdaFGL, a decoupled two-step personalized approach. Firstly, AdaFGL aggregates uploaded models to obtain the federated knowledge extractor through collaborative training. Then, each client conducts personalized training based on local subgraphs and the federated knowledge extractor. Experimental results on 12 graph benchmark datasets demonstrate the superior performance of AdaFGL over state-of-the-art baselines, achieving significant accuracy gains. |
| Low | GrooveSquid.com (original content) | Low Difficulty Summary Imagine a way to share information between different groups without sharing their individual data. This is what Federated Graph Learning (FGL) does! Currently, researchers use a method called community split to simulate this process. However, in real life, the groups might have different connections and structures, making it harder for FGL to work well. To solve this problem, we introduce a new approach called AdaFGL. It works by first sharing information among all groups to create a common understanding, then allowing each group to use its own local data to make decisions. Our results show that AdaFGL performs much better than previous methods in different scenarios. |
Keywords
* Artificial intelligence * Attention * Classification * Semi supervised
