Reproduce “Towards Explaining Query Expansion Performance in Information Retrieval” 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
Sourav Saha, Aditya Dutta, Soumajit Pramanik, Mandar Mitra
The authors present a post hoc diagnostic framework for query expansion that compares expanded queries with relevance-informed ideal queries and measures separation between judged-relevant and judged-non-relevant documents under BM25. On TREC Robust, Cohen’s d has a Pearson correlation of 0.7849 with average precision. The ideal queries use complete relevance judgments, so they are retrospective references, not deployable, label-free targets.
Summary by alphaXiv.
Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.