Summary of The Nah Bandit: Modeling User Non-compliance in Recommendation Systems, by Tianyue Zhou et al.
The Nah Bandit: Modeling User Non-compliance in Recommendation Systems
by Tianyue Zhou, Jung-Hoon Cho, Cathy Wu
First submitted to arxiv on: 15 Aug 2024
Categories
- Main: Machine Learning (cs.LG)
- Secondary: Information Retrieval (cs.IR); Multiagent Systems (cs.MA); Systems and Control (eess.SY)
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 Recommendation systems are ubiquitous online, but implementing them effectively in physical spaces like mobility or health remains challenging. This paper addresses a key issue: users can easily opt out of recommendations and fall back to their baseline behavior. To overcome this, the authors introduce the “Nah Bandit” problem, where users can reject recommendations and choose their preferred option instead. The authors model user non-compliance by parameterizing an anchoring effect of recommendations on users. They propose the Expert with Clustering (EWC) algorithm, a hierarchical approach that incorporates feedback from both recommended and non-recommended options to accelerate user preference learning. EWC achieves a regret bound of O(N√TlogK + NT) and outperforms traditional contextual bandit approaches in experimental results. This advancement reveals the importance of using non-compliance feedback to improve recommendation accuracy. |
| Low | GrooveSquid.com (original content) | Low Difficulty Summary Imagine trying to get people to use public transportation or exercise more often. It’s hard because they can just ignore your recommendations and do what they want instead. Researchers have been working on a way to make recommendations work better in these situations. They came up with something called the “Nah Bandit” problem, where people can say no to a recommendation and choose something else. The team found that by understanding how people react to recommendations, they could come up with a new algorithm that makes personalized suggestions more effective. This is important because it could help people make better choices about things like transportation or health. |
Keywords
» Artificial intelligence » Clustering
