Reproduce “Preference-based learning for news headline recommendation” on OpenResearch
Turn your favorite coding agents into research agents that review literature, develop hypotheses, and run experiments.
Universal app for Apple silicon and Intel. Requires macOS 11 or later.
About this paper
Alexandre Bouras, Audrey Durand, Richard Khoury
The paper learns preferences among available social-news headlines from pairwise engagement histories rather than predicting exact click counts. In a historical simulation using Facebook posts from one newspaper, greedy and NeuralTS both outperformed random choice in cumulative clicks, and neither surpassed the best-choice oracle. Greedy showed no observed disadvantage versus NeuralTS, but this was not a live test.
Summary by alphaXiv.
Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.