About Me
I am an engineer at Tencent since 2017, focusing on Recommendation Algorithms and Reinforcement Learning.
Our team has extensive experience in Recommendation System (RS) and has proposed many innovative solutions, achieving
significant benefits. Some of our solutions have been made public and successfully deployed in many companies' large-scale
Recommender Systems (RSs). The ideas behind our public solutions have also inspired related work in the industry.
We have a deep understanding of RSs and firmly believe that only solutions proven effective in real-world RSs are truly
valuable, rather than those that merely appear complex or grandiose. We welcome communication with professionals via email.
My ORCID: https://orcid.org/0009-0000-7271-4721
Our team has extensive experience in Recommendation System (RS) and has proposed many innovative solutions, achieving
significant benefits. Some of our solutions have been made public and successfully deployed in many companies' large-scale
Recommender Systems (RSs). The ideas behind our public solutions have also inspired related work in the industry.
We have a deep understanding of RSs and firmly believe that only solutions proven effective in real-world RSs are truly
valuable, rather than those that merely appear complex or grandiose. We welcome communication with professionals via email.
My ORCID: https://orcid.org/0009-0000-7271-4721
Experience
Tencent, 06/2017---Now
recommendation algorithm engineer
Our Publicly Released Solutions
-
Dynamic User Interest Augmentation via Stream Clustering and Memory Networks in Large-Scale Recommender Systems
https://arxiv.org/abs/2405.13238
Peng Liu, Nian Wang, Cong Xu, Ming Zhao, Bin Wang, Yi Ren.
This work was proposed and has been deployed in multiple large-scale RSs at Tencent since 2022.
It was make public on Tue, 21 May 2024. This solution has also been launched in many large-scale RSs and
inspired related works.
-
UnifiedRL: A Reinforcement Learning Algorithm Tailored for Multi-Task Fusion in Large-Scale Recommender Systems
https://arxiv.org/abs/2404.17589
Peng Liu, Cong Xu, Ming Zhao, Jiawei Zhu, Bin Wang, Yi Ren.
This work was proposed in 2023 and has been deployed in several large-scale RSs at Tencent since July 2023.
It was made public on Friday, April 19, 2024, which has been adopted in multiple large-scale RSs.
-
EnhancedRL: An Enhanced-State Reinforcement Learning Algorithm for Multi-Task Fusion in Recommender Systems
https://arxiv.org/abs/2409.11678
Peng Liu, Cong Xu, Jiawei Zhu, Ming Zhao, Bin Wang.
This work was proposed in 2023 and has been deployed in large-scale RSs at Tencent since September 14, 2023.
It was made public on Wed, September 18, 2024. This solution breaks through the traditional RL-MTF modeling pattern.
Last update: 2025-04-20