解决大规模动态异构图嵌入的冷启动难题，实现高效实时更新。
Towards Practical Large-scale Dynamical Heterogeneous Graph Embedding: Cold-start Resilient Recommendation
- 分两阶段设计：先用图变压器学全局结构，再用轻量算法实时增量更新。
- 在十亿级图上，广告价值提升达6.11%，刷新速度提高83.2%。
- 适合需要低延迟、高可扩展性的工业级推荐系统部署。
将动态异构图嵌入部署于生产环境面临可扩展性、数据新鲜度和冷启动等关键挑战。本文提出一种实用的两阶段解决方案，兼顾深度图表示与低延迟增量更新。框架结合 HetSGFormer（一种具有线性可扩展性的静态图变压器）与 Incremental Locally Linear Embedding（ILLE，一种轻量级、基于CPU的实时更新算法）。HetSGFormer 捕获全局结构，而 ILLE 实现快速、精准的增量更新，避免全量重训练。该双阶段方法具备冷启动鲁棒性，能利用图结构从稀疏数据中生成有意义的嵌入。在十亿规模图上，A/B测试显示，HetSGFormer 相较以往方法带来最高6.11%的广告商价值提升，而 ILLE 模块额外贡献3.22%提升，并使嵌入刷新时效性提高83.2%。本工作为动态图学习在生产环境中的部署提供了验证框架。
原文摘要 · Abstract (English)
Deploying dynamic heterogeneous graph embeddings in production faces key challenges of scalability, data freshness, and cold-start. This paper introduces a practical, two-stage solution that balances deep graph representation with low-latency incremental updates. Our framework combines HetSGFormer, a scalable graph transformer for static learning, with Incremental Locally Linear Embedding (ILLE), a lightweight, CPU-based algorithm for real-time updates. HetSGFormer captures global structure with linear scalability, while ILLE provides rapid, targeted updates to incorporate new data, thus avoiding costly full retraining. This dual approach is cold-start resilient, leveraging the graph to create meaningful embeddings from sparse data. On billion-scale graphs, A/B tests show HetSGFormer achieved up to a 6.11% lift in Advertiser Value over previous methods, while the ILLE module added another 3.22% lift and improved embedding refresh timeliness by 83.2%. Our work provides a validated framework for deploying dynamic graph learning in production environments.
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。