用语义融合提升生成式推荐，更好理解商品关系和长文本信息
GRAM: Generative Recommendation via Semantic-aware Multi-granular Late Fusion
- 将商品隐含关系转化为语言模型可懂的词汇编码
- 多粒度信息延迟融合，减少长文本信息损失
- 在4个数据集上超越主流模型，最高提升16%召回率
生成式推荐将推荐任务转化为文本生成，利用大语言模型的广泛知识。但现有方法存在两个问题：难以融入隐含的商品间关系，且对冗长的商品信息利用效率低。为此，我们提出语义感知的多粒度延迟融合生成推荐模型GRAM。首先，设计语义到词汇的转换机制，将商品的层级与协作关系编码进语言模型的词空间；其次，采用多粒度延迟融合策略，对不同粒度的提示分别编码，并推迟融合至解码阶段，实现高效信息整合与最小信息丢失。在四个基准数据集上的实验表明，GRAM优于8种前沿生成式推荐模型，在Recall@5上提升11.5-16.0%，NDCG@5上提升5.3-13.6%。代码已开源。
原文摘要 · Abstract (English)
Generative recommendation is an emerging paradigm that leverages the extensive knowledge of large language models by formulating recommendations into a text-to-text generation task. However, existing studies face two key limitations in (i) incorporating implicit item relationships and (ii) utilizing rich yet lengthy item information. To address these challenges, we propose a Generative Recommender via semantic-Aware Multi-granular late fusion (GRAM), introducing two synergistic innovations. First, we design semantic-to-lexical translation to encode implicit hierarchical and collaborative item relationships into the vocabulary space of LLMs. Second, we present multi-granular late fusion to integrate rich semantics efficiently with minimal information loss. It employs separate encoders for multi-granular prompts, delaying the fusion until the decoding stage. Experiments on four benchmark datasets show that GRAM outperforms eight state-of-the-art generative recommendation models, achieving significant improvements of 11.5-16.0% in Recall@5 and 5.3-13.6% in NDCG@5. The source code is available at https://github.com/skleee/GRAM.
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。