Reproduce “Graph Contrastive Learning with Multi-Objective for Personalized Product Retrieval in Taobao Search” 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
Longbin Li, Chao Zhang, Sen Li, Yun Zhong, Qingwen Liu, Xiaoyi Zeng
Graph Contrastive Learning with Multi-Objective (GCL-MO) integrates graph neural networks, multi-objective optimization, and contrastive learning to enhance personalized product retrieval in e-commerce. Successfully deployed in Taobao's search system, this framework improved Gross Merchandise Volume and increased engagement with long-tail items by learning richer, more balanced item representations.
Key insights
- Combining graph neural networks, multi-objective optimization, and contrastive learning effectively addresses multiple challenges in e-commerce retrieval.
- Multi-objective optimization yields more comprehensive item representations by balancing various user behaviors.
- Contrastive learning significantly improves the quality and discoverability of long-tail item representations.
Summary by alphaXiv.
Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.