用大模型做用户偏好先验，融合图文信息提升推荐精准度
RecMind: LLM-Enhanced Graph Neural Networks for Personalized Consumer Recommendations
- 将冻结大模型作为偏好先验，生成文本条件嵌入
- 在Yelp和Amazon上相对基线提升最高4.53%（Recall@40）
- 适合冷启动、长尾商品推荐场景
个性化推荐是消费科技、流媒体、购物、可穿戴设备和语音系统的核心能力，但面临交互稀疏、内容更新快、文本信号异构等挑战。我们提出RecMind，一种基于大语言模型增强的图神经网络推荐系统，将语言模型视为偏好先验而非单一排序器。一个配备轻量适配器的冻结大模型，从标题、属性和评论中生成文本条件用户/物品嵌入；轻量级LightGCN主干则从用户-物品图中学习协同嵌入。通过对称对比目标对齐两种视图，并采用层内门控机制融合，使语言信号在冷启动和长尾场景主导，图结构在其他场景稳定排名。在Yelp和Amazon-Electronics数据集上，RecMind在全部八个指标上表现最佳，相对强基线最高提升4.53%（Recall@40）和4.01%（NDCG@40）。消融实验验证了跨视图对齐的必要性以及门控融合优于后期融合和仅使用大模型的方案。
原文摘要 · Abstract (English)
Personalization is a core capability across consumer technologies, streaming, shopping, wearables, and voice, yet it remains challenged by sparse interactions, fast content churn, and heterogeneous textual signals. We present RecMind, an LLM-enhanced graph recommender that treats the language model as a preference prior rather than a monolithic ranker. A frozen LLM equipped with lightweight adapters produces text-conditioned user/item embeddings from titles, attributes, and reviews; a LightGCN backbone learns collaborative embeddings from the user-item graph. We align the two views with a symmetric contrastive objective and fuse them via intra-layer gating, allowing language to dominate in cold/long-tail regimes and graph structure to stabilize rankings elsewhere. On Yelp and Amazon-Electronics, RecMind attains the best results on all eight reported metrics, with relative improvements up to +4.53\% (Recall@40) and +4.01\% (NDCG@40) over strong baselines. Ablations confirm both the necessity of cross-view alignment and the advantage of gating over late fusion and LLM-only variants.
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。