Summary of Be Aware Of the Neighborhood Effect: Modeling Selection Bias Under Interference, by Haoxuan Li et al.
Be Aware of the Neighborhood Effect: Modeling Selection Bias under Interference
by Haoxuan Li, Chunyuan Zheng, Sihao Ding, Peng Wu, Zhi Geng, Fuli Feng, Xiangnan He
First submitted to arxiv on: 30 Apr 2024
Categories
- Main: Machine Learning (cs.LG)
- Secondary: Information Retrieval (cs.IR); Machine Learning (stat.ML)
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 addresses selection bias in recommender systems by formalizing the “neighborhood effect” as an interference problem from a causal inference perspective. This builds upon previous work that focused on addressing selection bias, but ignored this crucial aspect. The authors introduce a treatment representation to capture the neighborhood effect and propose a novel ideal loss function to deal with selection bias in its presence. They also develop two new estimators for estimating this ideal loss. The paper theoretically establishes connections between proposed and existing debiasing methods, showing that their approach can achieve unbiased learning when both selection bias and neighborhood effect are present. This is demonstrated through extensive semi-synthetic and real-world experiments. |
| Low | GrooveSquid.com (original content) | Low Difficulty Summary This paper helps to make online recommendations fairer by fixing a problem called “selection bias.” Selection bias happens when the system making the recommendation favors certain things over others, and users interact with these things in different ways. Researchers have tried to fix this problem before, but they haven’t considered that what one user chooses might affect what another user chooses. The authors of this paper come up with a new way to deal with both selection bias and this “neighborhood effect.” They show that their method can make recommendations fairer than previous methods. |
Keywords
» Artificial intelligence » Inference » Loss function
