---
格式版本: 2
标题: "FedCGR: Federated Cross-Domain Generative Recommendation"
原文链接: "https://arxiv.org/abs/2608.10929"
发布日期: "2026-08-11"
发布时间校准状态: "found"
发布时间需复核: "否"
发布时间来源: "rule:local:strict_original_body"
发布时间证据: "**\\[v1\\]** Tue, 11 Aug 2026 13:58:55 UTC (859 KB)"
发布时间校准原因: "规则确认唯一严格发布时间，来源 local:strict_original_body"
发布时间校准置信度: "high"
发布时间候选数量: 18
发布时间严格候选数量: 6
发布时间原页读取状态: "source template page reused from URL open"
发布时间未找到原因: ""
发布时间校准时间: "2026-08-12T11:15:47+08:00"
发布时间仲裁状态: "skipped"
发布时间仲裁尝试次数: 0
发布时间仲裁耗时毫秒: 0
发现时间: "2026-08-12T11:13:54+08:00"
入库时间: "2026-08-12T03:15:48.013Z"
来源平台: "arXiv 学术论文搜索"
搜索渠道: "source_template"
搜索词: "https://arxiv.org/search/?query=deployment&searchtype=all"
匹配关键词:
  - "deployment"
  - "performance"
  - "AI"
相关厂家:
  []
相关专家:
  []
内容类型: "网页"
抓取工具: "Free Fetch + Defuddle"
清洗工具: "Defuddle Markdown + Defuddle/Readability 正文提取"
原始附件:
  []
AI优质: "否"
AI打分: 5
AI分档: "非优质"
AI质检状态: "不通过"
AI打分理由: "这是一篇关于联邦跨域生成推荐的学术论文，与超节点/AI Rack/机柜级AI基础设施无关。"
AI质检模型: "ali-deepseek-v4-flash"
AI质检时间: "2026-08-12T11:15:52+08:00"
AI主题相关性: 0
AI来源权威性: 5
AI新颖性: 0
AI技术细节: 0
AI商业部署信号: 0
AI完整性: 0
采集批次: "2026年8月12日2点04分20秒"
采集批次ID: "20260812-020420-289"
去重键: "https://arxiv.org/abs/2608.10929"
---

## Computer Science > Artificial Intelligence

## Title:FedCGR: Federated Cross-Domain Generative Recommendation

Authors:[Zhuodong Liu](https://arxiv.org/search/cs?searchtype=author&query=Liu,+Z), [Hugen Lv](https://arxiv.org/search/cs?searchtype=author&query=Lv,+H), [Xiangyu Li](https://arxiv.org/search/cs?searchtype=author&query=Li,+X), [Bohan Guo](https://arxiv.org/search/cs?searchtype=author&query=Guo,+B), [Peiyu Hu](https://arxiv.org/search/cs?searchtype=author&query=Hu,+P)

[View PDF](https://arxiv.org/pdf/2608.10929) [HTML (experimental)](https://arxiv.org/html/2608.10929v1)

> Abstract:Cross-domain recommendation (CDR) transfers preference knowledge across related domains, but federated deployment makes cross-domain alignment difficult because the behavioral anchors that align item spaces, such as overlapping users and shared interaction signals, are often sparse, unavailable, or privacy-sensitive across clients. To address this tension, we revisit federated CDR as generation over a stable semantic item language. By representing items as discrete semantic ID (SID) sequences derived from public item-side metadata, cross-domain item alignment is induced by a shared vocabulary rather than by exchanging private interactions or aligning domain-specific embeddings. Directly federating SID-based generators, however, introduces two design constraints: the SID tokenizer must remain fixed to preserve cross-client token consistency, which creates a semantic-only bottleneck because local collaborative filtering (CF) signals cannot be globally shared or aligned; meanwhile, standard federated averaging can cause negative transfer under domain heterogeneity. To overcome these constraints, we propose FedCGR, a federated generative CDR framework that keeps the item language stable and makes adaptation explicit. FedCGR injects local CF evidence through a reliability-aware semantic interface and trains a prototype-personalized generator that selectively aggregates shared parameters according to domain relatedness while keeping domain-specific quantities local. Experiments on six Amazon cross-domain scenarios show that FedCGR consistently outperforms federated generative baselines and achieves competitive performance against strong sequential and federated CDR methods under both full-ranking and sampled evaluation protocols.

| Comments: |  |
| --- | --- |
| Subjects: | Artificial Intelligence (cs.AI) |
| Cite as: | [arXiv:2608.10929](https://arxiv.org/abs/2608.10929) \[cs.AI\] |
|  | (or [arXiv:2608.10929v1](https://arxiv.org/abs/2608.10929v1) \[cs.AI\] for this version) |
|  | [https://doi.org/10.48550/arXiv.2608.10929](https://doi.org/10.48550/arXiv.2608.10929) |

## Submission history

From: Zhuodong Liu \[[view email](https://arxiv.org/show-email/e3558bfe/2608.10929)\]  
**\[v1\]** Tue, 11 Aug 2026 13:58:55 UTC (859 KB)

[Which authors of this paper are endorsers?](https://arxiv.org/auth/show-endorsers/2608.10929) | Disable MathJax ([What is MathJax?](https://info.arxiv.org/help/mathjax.html))