Summary of Towards Robust Recommender Systems Via Triple Cooperative Defense, by Qingyang Wang et al.
Towards Robust Recommender Systems via Triple Cooperative Defense
by Qingyang Wang, Defu Lian, Chenwang Wu, Enhong Chen
First submitted to arxiv on: 25 Oct 2022
Categories
- Main: Machine Learning (cs.LG)
- Secondary: Cryptography and Security (cs.CR); 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 proposed Triple Cooperative Defense (TCD) framework aims to improve the robustness of recommender systems against well-crafted fake profiles by integrating data processing and robust modeling. The method co-trains three models, using high-confidence prediction ratings as auxiliary training data to enhance recommendation robustness. Unlike existing methods that either exclude normal samples or struggle with generalization and robustness, TCD avoids deleting abnormal data while improving model generalization through cooperative training. Experimental results on five poisoning attacks across three real-world datasets demonstrate significant robustness improvement compared to baselines. |
| Low | GrooveSquid.com (original content) | Low Difficulty Summary Recommender systems can be fooled by fake profiles, making it important to study how to defend against these attacks. The usual methods either remove normal samples or struggle with being both general and robust. A new approach called Triple Cooperative Defense (TCD) combines data processing and robust modeling to make recommendations better. Instead of removing bad data, TCD adds extra information to help the models learn. This makes the models not just more robust but also better at making predictions in general. In experiments with five types of attacks on three real-world datasets, TCD showed significant improvement over other methods. |
Keywords
» Artificial intelligence » Generalization
