Sequential Recommendation via Adaptive Robust Attention with Multi-dimensional Embeddings
Fuente:
arXiv
Gespeichert in:
| Hauptverfasser: | , , , |
|---|---|
| Format: | Preprint |
| Veröffentlicht: |
2024
|
| Schlagworte: | |
| Online-Zugang: | |
| Tags: |
Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
|
| _version_ | 1866910594827812864 |
|---|---|
| author | Pang, Linsey Raffiee, Amir Hossein Liu, Wei Lundgaard, Keld |
| author_facet | Pang, Linsey Raffiee, Amir Hossein Liu, Wei Lundgaard, Keld |
| contents | Sequential recommendation models have achieved state-of-the-art performance using self-attention mechanism. It has since been found that moving beyond only using item ID and positional embeddings leads to a significant accuracy boost when predicting the next item. In recent literature, it was reported that a multi-dimensional kernel embedding with temporal contextual kernels to capture users' diverse behavioral patterns results in a substantial performance improvement. In this study, we further improve the sequential recommender model's robustness and generalization by introducing a mix-attention mechanism with a layer-wise noise injection (LNI) regularization. We refer to our proposed model as adaptive robust sequential recommendation framework (ADRRec), and demonstrate through extensive experiments that our model outperforms existing self-attention architectures. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2409_05022 |
| institution | arXiv |
| publishDate | 2024 |
| record_format | arxiv |
| spellingShingle | Sequential Recommendation via Adaptive Robust Attention with Multi-dimensional Embeddings Pang, Linsey Raffiee, Amir Hossein Liu, Wei Lundgaard, Keld Information Retrieval Artificial Intelligence Machine Learning Sequential recommendation models have achieved state-of-the-art performance using self-attention mechanism. It has since been found that moving beyond only using item ID and positional embeddings leads to a significant accuracy boost when predicting the next item. In recent literature, it was reported that a multi-dimensional kernel embedding with temporal contextual kernels to capture users' diverse behavioral patterns results in a substantial performance improvement. In this study, we further improve the sequential recommender model's robustness and generalization by introducing a mix-attention mechanism with a layer-wise noise injection (LNI) regularization. We refer to our proposed model as adaptive robust sequential recommendation framework (ADRRec), and demonstrate through extensive experiments that our model outperforms existing self-attention architectures. |
| title | Sequential Recommendation via Adaptive Robust Attention with Multi-dimensional Embeddings |
| topic | Information Retrieval Artificial Intelligence Machine Learning |
| url | https://arxiv.org/abs/2409.05022 |