recommendation systems 2bias correction 1causal inference 1deep learning 1e-commerce 1heterogeneous content 1large-scale systems 1modular architecture 1offline evaluation 1retrieval scaling 1
From the 2 of 3 linked papers with an AI index.
3 papers
cs.IR2026
PinEqualizer: Full Funnel Content Exploration and Debiasing System at Pinterest
Olafur Gudmundsson, Bo Zhao, Huayi Liao +15
In this paper, we propose a new solution for addressing the content cold-start problem in industry-scale search and recommender systems. Compared to prior approaches, we have made…
cs.IR2026
Deep-learning Causal Retrieval Optimization for Efficient e-commerce Distribution in Pinterest
Junpeng Hou, XianXing Zhang, Sai Xiao +6
The paper presents a deep learning system that decides when to trigger shopping recommendations in Pinterest, using causal inference to reduce unnecessary triggers while maintainin…
cs.IR2026
MESH: Scaling Up Retrieval with Heterogeneous Content Unification
Jiaxing Qu, Yilin Chen, Junpeng Hou +4
The paper introduces MESH, a unified framework that reduces bias in heterogeneous large‑scale retrieval systems by modularizing feature spaces and applying gated bias correction, l…