GRAM: Generative Recommendation via Semantic-aware Multi-granular Late Fusion
Fuente:
arXiv
Gespeichert in:
| Hauptverfasser: | , , , , |
|---|---|
| Format: | Preprint |
| Veröffentlicht: |
2025
|
| Schlagworte: | |
| Online-Zugang: | |
| Tags: |
Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
|
| _version_ | 1866918042938638336 |
|---|---|
| author | Lee, Sunkyung Choi, Minjin Choi, Eunseong Kim, Hye-young Lee, Jongwuk |
| author_facet | Lee, Sunkyung Choi, Minjin Choi, Eunseong Kim, Hye-young Lee, Jongwuk |
| contents | 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. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2506_01673 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | GRAM: Generative Recommendation via Semantic-aware Multi-granular Late Fusion Lee, Sunkyung Choi, Minjin Choi, Eunseong Kim, Hye-young Lee, Jongwuk Information Retrieval Artificial Intelligence Computation and Language 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. |
| title | GRAM: Generative Recommendation via Semantic-aware Multi-granular Late Fusion |
| topic | Information Retrieval Artificial Intelligence Computation and Language |
| url | https://arxiv.org/abs/2506.01673 |