Yihao Wang
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Previous Organizations / Institutes
The organizations / institutes from the scholar's other articles in the past ten years
Yihao Wang has a total of 26 co-authors. Here are the top co-authors by number of shared publications:
- Enqiang Xu
JD.com, Beijing, China
2 shared publications - Jinyuan Zhao
JD.com, Beijing, China
2 shared publications - Songlin Wang
JD.com, Beijing, China
2 shared publications Fangcheng Zhang
Department of Neurosurgery, Union Hospital, Tongji Medical College, Huazhong University of Science and Technology, Wuhan, 430022, People’s Republic of China
1 shared publication- Guangtao Nie
JD.com, Beijing, China
1 shared publication - Guoyu Tang
JD.com, Beijing, China
1 shared publication
Latest Publications
- Open Access
- 19 July 2026
- Published by Association for Computing Machinery (ACM)
- p. 4955-4959
Query-Attention Dual-Stream Framework with Cross-Category Transfer for Efficient Fine-Grained Interest Pre-Ranking
Large-scale search and recommendation systems typically adopt a cascaded architecture of retrieval, pre-ranking, ranking, and re-ranking to balance efficiency and accuracy. However, pre-ranking still faces challenges of behavioral sparsity, limited interest diversity, and computational latency. We propose the Query-Attention Dual-Stream (QADS) framework to address these issues. QADS partitions user behaviors into strongly and weakly correlated streams and further decomposes them into fine-grained subsequences guided by domain knowledge. A query-centric attention mechanism reduces complexity from O(N) to O(1), enabling efficient inter- and intra-sequence modeling. A contrastive cross-category transfer module propagates dense patterns from weakly correlated to sparse domains, while a latency-aware parallel inference architecture further reduces delay by 36%. Experiments on public and industrial datasets show that QADS delivers significant performance improvements and has been successfully deployed in large-scale e-commerce search systems.
- 19 July 2026
- Published by Association for Computing Machinery (ACM)
- p. 4955-4959
Open AccessQuery-Attention Dual-Stream Framework with Cross-Category Transfer for Efficient Fine-Grained Interest Pre-Ranking
Large-scale search and recommendation systems typically adopt a cascaded architecture of retrieval, pre-ranking, ranking, and re-ranking to balance efficiency and accuracy. However, pre-ranking still faces challenges of behavioral sparsity, limited interest diversity, and computational latency. We propose the Query-Attention Dual-Stream (QADS) framework to address these issues. QADS partitions user behaviors into strongly and weakly correlated streams and further decomposes them into fine-grained subsequences guided by domain knowledge. A query-centric attention mechanism reduces complexity from O(N) to O(1), enabling efficient inter- and intra-sequence modeling. A contrastive cross-category transfer module propagates dense patterns from weakly correlated to sparse domains, while a latency-aware parallel inference architecture further reduces delay by 36%. Experiments on public and industrial datasets show that QADS delivers significant performance improvements and has been successfully deployed in large-scale e-commerce search systems.
- Open Access
- 21 October 2024
- Published by Association for Computing Machinery (ACM)
- p. 3694-3698
MODRL-TA: A Multi-Objective Deep Reinforcement Learning Framework for Traffic Allocation in E-Commerce Search
Traffic allocation is a process of redistributing natural traffic to products by adjusting their positions in the post-search phase, aimed at effectively fostering merchant growth, precisely meeting customer demands, and ensuring the maximization of interests across various parties within e-commerce platforms. Existing methods based on learning to rank neglect the long-term value of traffic allocation, whereas approaches of reinforcement learning suffer from balancing multiple objectives and the difficulties of cold starts within real-world data environments. To address the aforementioned issues, this paper propose a multi-objective deep reinforcement learning framework consisting of multi-objective Q-learning (MOQ), a decision fusion algorithm (DFM) based on the cross-entropy method(CEM), and a progressive data augmentation system (PDA). Specifically. MOQ constructs ensemble RL models, each dedicated to an objective, such as click-through rate, conversion rate, etc. These models individually determine the position of items as actions, aiming to estimate the long-term value of multiple objectives from an individual perspective. Then we employ DFM to dynamically adjust weights among objectives to maximize long-term value, addressing temporal dynamics in objective preferences in e-commerce scenarios. Initially, PDA trained MOQ with simulated data from offline logs. As experiments progressed, it strategically integrated real user interaction data, ultimately replacing the simulated dataset to alleviate distributional shifts and the cold start problem. Experimental results on real-world online e-commerce systems demonstrate the significant improvements of MODRL-TA, and we have successfully deployed MODRL-TA on an e-commerce search platform.
- 21 October 2024
- Published by Association for Computing Machinery (ACM)
- p. 3694-3698
Open AccessMODRL-TA: A Multi-Objective Deep Reinforcement Learning Framework for Traffic Allocation in E-Commerce Search
Traffic allocation is a process of redistributing natural traffic to products by adjusting their positions in the post-search phase, aimed at effectively fostering merchant growth, precisely meeting customer demands, and ensuring the maximization of interests across various parties within e-commerce platforms. Existing methods based on learning to rank neglect the long-term value of traffic allocation, whereas approaches of reinforcement learning suffer from balancing multiple objectives and the difficulties of cold starts within real-world data environments. To address the aforementioned issues, this paper propose a multi-objective deep reinforcement learning framework consisting of multi-objective Q-learning (MOQ), a decision fusion algorithm (DFM) based on the cross-entropy method(CEM), and a progressive data augmentation system (PDA). Specifically. MOQ constructs ensemble RL models, each dedicated to an objective, such as click-through rate, conversion rate, etc. These models individually determine the position of items as actions, aiming to estimate the long-term value of multiple objectives from an individual perspective. Then we employ DFM to dynamically adjust weights among objectives to maximize long-term value, addressing temporal dynamics in objective preferences in e-commerce scenarios. Initially, PDA trained MOQ with simulated data from offline logs. As experiments progressed, it strategically integrated real user interaction data, ultimately replacing the simulated dataset to alleviate distributional shifts and the cold start problem. Experimental results on real-world online e-commerce systems demonstrate the significant improvements of MODRL-TA, and we have successfully deployed MODRL-TA on an e-commerce search platform.
- Open Access
- 1 May 2021
- Published by Taylor & Francis in Cancer Management and Research
- Vol. ume 13, 4217-4232
Remodelling and Treatment of the Blood-Brain Barrier in Glioma
The blood-brain barrier (BBB) is an essential structure of the central nervous system (CNS), and its existence makes the local internal environment of the CNS a relatively independent structure distinct from other internal environments of the human body to ensure normal physiological and high stability of activities of the CNS. Changes in BBB structure and function are fundamental to the pathophysiology of many diseases. The occurrence and development of glioma are often accompanied by a series of changes in the structure and function of the internal environment, the most significant of which is remodelling of the BBB. The remodelling of the BBB usually leads to changes in the permeability of local microvessels, which provide certain favourable conditions for the occurrence and development of glioma. Meanwhile, the newly generated abnormal blood vessels and the remaining intact regions of the BBB also hinder the effects of drug treatments. Changes in permeability and structural function often lead to the creation of abnormally functioning vascular regions, which pose further treatment challenges. At present, therapeutic methods for glioma have not achieved satisfactory effects in clinical practice, and emerging therapeutic methods have not yet been widely used in clinical practice. In this review, we summarize the knowledge of remodelling of the BBB in the glioma environment, the type of changes that occur, and current BBB treatment methods and prospects for the treatment of glioma.
- 1 May 2021
- Published by Taylor & Francis in Cancer Management and Research
- Vol. ume 13, 4217-4232
Open AccessRemodelling and Treatment of the Blood-Brain Barrier in Glioma
The blood-brain barrier (BBB) is an essential structure of the central nervous system (CNS), and its existence makes the local internal environment of the CNS a relatively independent structure distinct from other internal environments of the human body to ensure normal physiological and high stability of activities of the CNS. Changes in BBB structure and function are fundamental to the pathophysiology of many diseases. The occurrence and development of glioma are often accompanied by a series of changes in the structure and function of the internal environment, the most significant of which is remodelling of the BBB. The remodelling of the BBB usually leads to changes in the permeability of local microvessels, which provide certain favourable conditions for the occurrence and development of glioma. Meanwhile, the newly generated abnormal blood vessels and the remaining intact regions of the BBB also hinder the effects of drug treatments. Changes in permeability and structural function often lead to the creation of abnormally functioning vascular regions, which pose further treatment challenges. At present, therapeutic methods for glioma have not achieved satisfactory effects in clinical practice, and emerging therapeutic methods have not yet been widely used in clinical practice. In this review, we summarize the knowledge of remodelling of the BBB in the glioma environment, the type of changes that occur, and current BBB treatment methods and prospects for the treatment of glioma.