Summary of Performance Evaluation Of Knowledge Graph Embedding Approaches Under Non-adversarial Attacks, by Sourabh Kapoor et al.
Performance Evaluation of Knowledge Graph Embedding Approaches under Non-adversarial Attacks
by Sourabh Kapoor, Arnab Sharma, Michael Röder, Caglar Demir, Axel-Cyrille Ngonga Ngomo
First submitted to arxiv on: 9 Jul 2024
Categories
- Main: Machine Learning (cs.LG)
- Secondary: Cryptography and Security (cs.CR)
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 knowledge graph embedding (KGE) paper tackles the critical issue of non-adversarial attacks on five state-of-the-art algorithms across three attack surfaces. The study evaluates the impact of label, parameter, and graph perturbations on KGE performance using five datasets. Results show that label perturbation has a significant effect, followed by parameter perturbation with a moderate effect, while graph perturbation has a low impact. This research provides crucial insights for improving the robustness of KGE approaches in AI-driven applications like semantic search, question answering, and recommenders. |
| Low | GrooveSquid.com (original content) | Low Difficulty Summary A group of researchers is working on a new way to understand big data. They’re looking at how we can make sure that this data isn’t being manipulated or changed in ways that affect its usefulness. Right now, some people are trying to attack this data by changing it slightly, which affects how computers use it. The scientists want to see how well different methods they’ve developed can handle these attacks. They tested five different approaches on five sets of data and found that one type of attack is particularly effective in making the data less useful. This research will help make sure that big data is used correctly in things like searching for information, answering questions, or giving recommendations. |
Keywords
* Artificial intelligence * Embedding * Knowledge graph * Question answering
