Reproduce “Fast and Feasible: Permutation-based Constrained Reranking for Revenue Maximization” 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
Svetlana Shirokovskikh, Anastasiia Soboleva, Ekaterina Solodneva, Aleksandr Katrutsa, Roman Loginov, Egor Samosvat
PermR reorders existing e-commerce search results to increase promoted-listing revenue while keeping modeled quality metrics at or above production-ranking baselines. It uses local swaps, guided by estimated metric contributions and per-query bounds, rather than requiring an exact optimization solution in the live path. In a 14-day test across about 56 million queries, income rose 2.0% ± 0.4%; modeled constraints were preserved.
Summary by alphaXiv.
Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.