Summary of Hound: Hunting Supervision Signals For Few and Zero Shot Node Classification on Text-attributed Graph, by Yuxiang Wang et al.
Hound: Hunting Supervision Signals for Few and Zero Shot Node Classification on Text-attributed Graph
by Yuxiang Wang, Xiao Yan, Shiyu Jin, Quanqing Xu, Chuanhui Yang, Yuanyuan Zhu, Chuang Hu, Bo Du, Jiawei Jiang
First submitted to arxiv on: 1 Sep 2024
Categories
- Main: Artificial Intelligence (cs.AI)
- Secondary: Computation and Language (cs.CL); Information Retrieval (cs.IR)
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 paper proposes a new method called Hound for improving few- and zero-shot node classification on text-attributed graphs (TAGs). TAGs have text descriptions for each node, which is useful in applications such as academia and social networks. Existing methods only use contrastive loss to align graph-based node embedding and language-based text embedding, but this paper introduces more supervision signals by designing three augmentation techniques: node perturbation, text matching, and semantics negation. These techniques provide more reference nodes for each text and vice versa, allowing for better node-text pairings. The authors evaluate Hound on 5 datasets and compare it to 13 state-of-the-art baselines, showing that it consistently outperforms all of them. |
| Low | GrooveSquid.com (original content) | Low Difficulty Summary This paper is about a new way to improve computers’ ability to understand graphs with text descriptions. Graphs are like maps, but instead of showing where things are located, they show how things are related. The problem is that there aren’t always enough labels or clues for the computer to figure out what each thing does or means. This paper proposes a new method called Hound that helps computers better understand graphs with text descriptions by providing more clues and helping them learn from similar texts and nodes. |
Keywords
» Artificial intelligence » Classification » Contrastive loss » Embedding » Semantics » Zero shot
