Kuaishou Academic Forum on Cutting-Edge Short Video Recommendation Algorithms (Nov 23, 2023)

In recent years, short videos have become a dominant media format due to fragmented user time, the rapid spread of mobile internet, and low production barriers; platforms such as Kuaishou and Douyin attract massive audiences, largely thanks to sophisticated recommendation algorithms.

Kuaishou’s research team has published over a hundred papers at top conferences (SIGIR, KDD, WWW) and received honors like SIGIR 2023 Best Paper Honorable Mention and CIKM 2022 Best Paper, highlighting the academic impact of its work.

On the afternoon of November 23, Kuaishou, together with the Chinese Information Retrieval Committee of the China Computer Federation, hosted the CIR‑Kuaishou Academic Forum titled “Frontier Technologies in Short‑Video Recommendation Algorithms.” The event took place at the Silver Ginkgo Hall, Building 1, 3rd Floor, West Suburban Hotel, Haidian District, Beijing.

Forum Schedule and Highlights

• 14:00‑14:05 – Opening remarks by Jiang Peng, Vice President of Kuaishou. • 14:05‑14:30 – Keynote by Prof. Zhang Min (Tsinghua University) – “Short‑Video Recommendation: Opportunities and Challenges of Interdisciplinary Research.”

Session 1 – Reinforcement Learning • 14:30‑14:50 – Cai Qingpeng, Senior Algorithm Expert, Kuaishou – “Reinforcement Learning for Short Video Recommender Systems.” • 14:50‑15:10 – Gao Chongming, PhD candidate, University of Science and Technology of China – “Alleviating the Matthew Effect of Offline Reinforcement Learning in Interactive Recommendation” (SIGIR 2023 Best Paper Honorable Mention). • 15:10‑15:30 – Liu Shuchang, Senior Algorithm Engineer, Kuaishou – “KuaiSim: A Comprehensive Simulator for Recommender Systems” (NeurIPS 2023).

Break (15:30‑15:45).

Session 2 – User Interest Modeling • 15:45‑16:05 – Wu Cheng, PhD candidate, Tsinghua University – “Graph Neural Networks for Personalized Recommendation.” • 16:05‑16:25 – Si Zihua, Master’s student, Renmin University – “When Search Meets Recommendation: Learning Disentangled Search Representation for Recommendation” (SIGIR 2023). • 16:25‑16:45 – Hui Yiqun, Senior Recommendation Algorithm Engineer, Kuaishou – “TWIN: Two‑Stage Interest Network for Lifelong User Behavior Modeling in CTR Prediction” (KDD 2023).

• 16:45‑17:00 – Closing remarks. • 17:30‑20:00 – Kuaishou Campus Recruitment Dinner.

Attendees were invited to register by scanning the QR code provided in the announcement.

Company Overview

Kuaishou is a leading content community and social platform dedicated to creating value for its users. It serves billions of users through short videos, live streaming, e‑commerce, online knowledge sharing, and local services, continuously innovating its products and technologies. Follow the “Kuaishou Technology Team” WeChat public account for the latest updates.