Future Impact Decomposition in Request-level Recommendations
Fuente:
arXiv
Saved in:
| Main Authors: | , , , , , , , , |
|---|---|
| Format: | Preprint |
| Published: |
2024
|
| Subjects: | |
| Online Access: | |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866929390752890880 |
|---|---|
| author | Wang, Xiaobei Liu, Shuchang Wang, Xueliang Cai, Qingpeng Hu, Lantao Li, Han Jiang, Peng Gai, Kun Xie, Guangming |
| author_facet | Wang, Xiaobei Liu, Shuchang Wang, Xueliang Cai, Qingpeng Hu, Lantao Li, Han Jiang, Peng Gai, Kun Xie, Guangming |
| contents | In recommender systems, reinforcement learning solutions have shown promising results in optimizing the interaction sequence between users and the system over the long-term performance. For practical reasons, the policy's actions are typically designed as recommending a list of items to handle users' frequent and continuous browsing requests more efficiently. In this list-wise recommendation scenario, the user state is updated upon every request in the corresponding MDP formulation. However, this request-level formulation is essentially inconsistent with the user's item-level behavior. In this study, we demonstrate that an item-level optimization approach can better utilize item characteristics and optimize the policy's performance even under the request-level MDP. We support this claim by comparing the performance of standard request-level methods with the proposed item-level actor-critic framework in both simulation and online experiments. Furthermore, we show that a reward-based future decomposition strategy can better express the item-wise future impact and improve the recommendation accuracy in the long term. To achieve a more thorough understanding of the decomposition strategy, we propose a model-based re-weighting framework with adversarial learning that further boost the performance and investigate its correlation with the reward-based strategy. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2401_16108 |
| institution | arXiv |
| publishDate | 2024 |
| record_format | arxiv |
| spellingShingle | Future Impact Decomposition in Request-level Recommendations Wang, Xiaobei Liu, Shuchang Wang, Xueliang Cai, Qingpeng Hu, Lantao Li, Han Jiang, Peng Gai, Kun Xie, Guangming Information Retrieval H.3.3 In recommender systems, reinforcement learning solutions have shown promising results in optimizing the interaction sequence between users and the system over the long-term performance. For practical reasons, the policy's actions are typically designed as recommending a list of items to handle users' frequent and continuous browsing requests more efficiently. In this list-wise recommendation scenario, the user state is updated upon every request in the corresponding MDP formulation. However, this request-level formulation is essentially inconsistent with the user's item-level behavior. In this study, we demonstrate that an item-level optimization approach can better utilize item characteristics and optimize the policy's performance even under the request-level MDP. We support this claim by comparing the performance of standard request-level methods with the proposed item-level actor-critic framework in both simulation and online experiments. Furthermore, we show that a reward-based future decomposition strategy can better express the item-wise future impact and improve the recommendation accuracy in the long term. To achieve a more thorough understanding of the decomposition strategy, we propose a model-based re-weighting framework with adversarial learning that further boost the performance and investigate its correlation with the reward-based strategy. |
| title | Future Impact Decomposition in Request-level Recommendations |
| topic | Information Retrieval H.3.3 |
| url | https://arxiv.org/abs/2401.16108 |