Summary of Path-based Explanation For Knowledge Graph Completion, by Heng Chang et al.
Path-based Explanation for Knowledge Graph Completion
by Heng Chang, Jiangnan Ye, Alejo Lopez Avila, Jinhua Du, Jia Li
First submitted to arxiv on: 4 Jan 2024
Categories
- Main: Machine Learning (cs.LG)
- Secondary: Artificial Intelligence (cs.AI); 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 The abstract proposes Power-Link, a novel path-based explainer for Knowledge Graph Completion (KGC) tasks, which leverages Graph Neural Networks (GNNs). The model aims to provide transparent explanations of GNN-based KGC results by exploring entity-relation interactions. It designs a simplified graph-powering technique and introduces new metrics for evaluating the generated explanations. Experimental results show that Power-Link outperforms state-of-the-art baselines in interpretability, efficiency, and scalability. |
| Low | GrooveSquid.com (original content) | Low Difficulty Summary Power-Link is a new way to explain how Graph Neural Networks work on big databases of knowledge. Right now, it’s hard to understand why these models make certain predictions. By looking at the paths between different things in the database, Power-Link helps make these predictions more transparent and reliable. The creators of this method came up with a simple but efficient way to train their model and developed new ways to measure how well it works. |
Keywords
* Artificial intelligence * Gnn * Knowledge graph
