Summary of Dirw: Path-aware Digraph Learning For Heterophily, by Daohan Su et al.
DiRW: Path-Aware Digraph Learning for Heterophily
by Daohan Su, Xunkai Li, Zhenjun Li, Yinping Liao, Rong-Hua Li, Guoren Wang
First submitted to arxiv on: 14 Oct 2024
Categories
- Main: Machine Learning (cs.LG)
- Secondary: Artificial Intelligence (cs.AI)
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 A novel graph neural network approach, Directed Random Walk (DiRW), is proposed to effectively learn from directed graphs. This paper addresses the limitations of existing DiGNNs and other spatial-based methods by providing a plug-and-play strategy or innovative architecture for directed graph learning. DiRW incorporates a direction-aware path sampler and node-wise learnable path aggregator to represent nodes in digraphs. Experimental results on 9 datasets demonstrate that DiRW enhances most spatial-based methods as a plug-and-play strategy, achieving state-of-the-art (SOTA) performance as a new digraph learning paradigm. |
| Low | GrooveSquid.com (original content) | Low Difficulty Summary Directed graphs are widely used in social networks and recommendations, offering a new perspective for addressing topological heterophily challenges. A new approach to learn from directed graphs is proposed, called Directed Random Walk (DiRW). DiRW provides a plug-and-play strategy or innovative architecture for most spatial-based methods or digraphs. This approach incorporates a direction-aware path sampler and node-wise learnable path aggregator to represent nodes in digraphs. The results show that DiRW achieves state-of-the-art performance on 9 datasets. |
Keywords
» Artificial intelligence » Graph neural network
