Reproduce “A Survey on User Behavior Modeling in Recommender Systems” 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

Zhicheng He, Weiwen Liu, Wei Guo, Jiarui Qin, Yingxue Zhang, Yaochen Hu, Ruiming Tang

This survey paper from Huawei Noah’s Ark Lab and Shanghai Jiao Tong University provides a comprehensive overview of recent advancements in User Behavior Modeling (UBM) for recommender systems. It proposes a novel taxonomy covering conventional, long-sequence, multi-type, and side information-enhanced UBM, detailing prevalent techniques and their practical application in large-scale industrial systems.

Key insights

  • UBM research is evolving towards modeling longer sequences, incorporating multiple behavior types, and integrating heterogeneous side information.
  • Computational efficiency and scalability are paramount for UBM deployment in large-scale industrial recommender systems, often leading to a gap between academic trends (e.g., GNNs) and practical adoption.
  • Effective UBM solutions require carefully designed fusion strategies for diverse data sources and a balance between model complexity and real-world deployability.

Summary by alphaXiv.

Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.