用电影推荐应用测试全同态加密部署可行性
Scaling Homomorphic Applications in Deployment
- 构建容器化电影推荐应用验证加密生态可用性
- 通过配置调优降低全同态加密计算开销
- 适合关注隐私计算落地的工程团队参考
本文开发了一个全同态加密（FHE）的原型应用，以评估加密生态系统的生产就绪状态。基于电影推荐场景，通过容器化和编排技术实现应用部署。通过对部署配置进行调优，有效缓解了全同态加密带来的计算瓶颈，展示了在实际环境中优化基础设施以支持加密计算的可行性。该工作为隐私保护型应用的规模化部署提供了实践路径。
原文摘要 · Abstract (English)
In this endeavor, a proof-of-concept homomorphic application is developed to determine the production readiness of encryption ecosystems. A movie recommendation app is implemented for this purpose and productionized through containerization and orchestration. By tuning deployment configurations, the computational limitations of Fully Homomorphic Encryption (FHE) are mitigated through additional infrastructure optimizations Index Terms: Reinforcement Learning, Orchestration, Homomorphic Encryption
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。