AI Paper Atlas

Research map / Learning theory and algorithms / Prediction and decision losses

Cache-Augmented Inbatch Importance Resampling for Training Recommender Retriever

Jin Chen, Defu Lian, Yucheng Li, Baoyun Wang, Kai Zheng, Enhong Chen

NeurIPS 2022 · Learning theory and algorithms › Prediction and decision losses · approach: Ranking models from comparisons · 4 citations (OpenAlex)

Read the paperCheck your idea against 137,217 papers

AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.

Measured effects reported in the paper

What was changedWhereMeasuredEffect
add item-popularity proposal correction to inbatch sampling (SSL-Pop vs SSL)SSL · GowallaNDCG@10: 0.1381 -> 0.1479Improved
add item-popularity proposal correction (SSL-Pop vs SSL)SSL · Ta-fengNDCG@10: 0.0437 -> 0.0625Improved
add uniformly sampled out-of-batch negatives (MNS vs SSL-Pop)SSL-Pop · EchonestNDCG@10: 0.1609 -> 0.1648Improved
replace inbatch popularity correction with importance resampling (BIR vs SSL-Pop)SSL-Pop · GowallaNDCG@10: 0.1479 -> 0.1523Improved
replace inbatch popularity correction with importance resampling (BIR vs best baseline)best baseline · Ta-fengNDCG@10: 0.0634 -> 0.0675Improved
add cache of frequently sampled items to BIR (XIR vs BIR) (ablation)BIR · AmazonNDCG@10: 0.0833 -> 0.0877Improved
add cache of frequently sampled items to BIR (XIR vs BIR) (ablation)BIR · EchonestNDCG@10: 0.1682 -> 0.1842Improved
add cache of frequently sampled items to BIR (XIR vs BIR) (ablation)BIR · GowallaRECALL@10: 0.1157 -> 0.1169Improved
add cache of frequently sampled items to BIR (XIR vs BIR) (ablation)BIR · TmallNDCG@10: 0.0590 -> 0.0658Improved
BIR vs SSL-Pop on Gowalla recall (G-Tower stronger)BIR · GowallaRECALL@10: 0.1176 vs 0.1157No clear effect
add item features (SellerID, CateID) with a two-layer item tower, BIR vs SSL-PopSSL-Pop · TmallNDCG@10: 3.06%Improved
add item features, XIR vs SSL-PopSSL-Pop · TmallNDCG@10: 4.87%Improved

Closest papers by meaning

Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy