解决生成推荐中冷启动物品编码僵化问题,动态优化初始标识分配。
DREAM: Dynamic Refinement of Early Assignment Mappings

- 引入三阶段框架,动态修正冷启动物品的初始标识分配。
- 在三个亚马逊数据集上,冷启动推荐效果显著优于当前最优方法。
- 适合关注冷启动推荐与生成式推荐融合的研究者。
生成式推荐通过将物品检索重构为语义标识(SIDs)的自回归生成来提升性能,其中SIDs是编码物品语义的紧凑标记序列。尽管SIDs提供强语义先验,现有方法在充分用户反馈前即通过离线分词为每个物品分配单一静态标识。对于冷启动物品,这种一次性承诺导致判别力弱的编码,产生错位路径且难以优化,因为相关标记在训练中极少被采样。我们发现,早期静态绑定而非模型容量,是基于SIDs的生成式推荐中冷启动的根本瓶颈。为此,提出DREAM(动态修正早期分配映射),通过渐进式精炼解决此缺陷:第一阶段,意图感知分词器通过反事实对比学习重建SID空间,为每个冷启动物品生成多样化的行为对齐候选;第二阶段,冻结的推荐主干作为评估器,基于多上下文用户支持选择最可靠候选,无需重训练;第三阶段,动态束机制在整个训练和推理过程中维护多个加权的SID假设,防止过早坍缩至单一分配。在三个Amazon基准上的大量实验表明,DREAM在冷启动指标上显著优于最先进的生成式与序列基线。
原文摘要 · Abstract (English)
Generative recommendation advances item retrieval by reformulating it as autoregressive generation of Semantic IDs (SIDs), compact token sequences that encode item semantics. While SIDs offer a strong semantic prior, current SID-based methods assign each item a single static identifier through offline tokenization before sufficient user feedback is observed. For cold-start items, this one-shot commitment produces poorly discriminative codes, generating misaligned paths that remain unrefined because the associated tokens are rarely sampled during training. We identify this early static commitment, not model capacity, as the fundamental cold-start bottleneck in SID-based generative recommendation. To overcome this bottleneck and bridge the disjoint objectives of tokenization and generation, we propose DREAM (Dynamic Refinement of Early Assignment Mappings), a three-stage framework that resolves this flaw through progressive refinement. First, an intent-aware tokenizer rebuilds the SID space through counterfactual contrastive learning, generating a diverse pool of behavior-aligned candidates per cold-start item. Second, the frozen recommendation backbone serves as an evaluator, selecting the most reliable candidate based on multi-context user support without retraining. Third, a dynamic beam mechanism maintains multiple weighted SID hypotheses throughout training and inference, preventing premature collapse to a single assignment. Extensive experiments on three Amazon benchmarks show that DREAM substantially outperforms state-of-the-art generative and sequential baselines on cold-start metrics.
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。