PhD Data Science, School of DS, City University of Hong kong, HongKong 999077, China Tel: +86 18810218320/+852 66618290 Email: langmiliu2-c@my.cityu.edu.hk, liulangming.llm@alibaba-inc.com [Google Scholar]

About Me

I am currently working in Taobao & Tmall Group of Alibaba as a senior algorithm engineer.

I earned my PhD in Data Science under the supervision of Prof. Xiangyu Zhao, with Prof. Dingxuan Zhou as my co-supervisor. I received my science bachelor's degree in 2020 from Beihang University, where I majored in Applied Mathematics.

My research interests include Large Language Models, Recommendation Systems, Federated Learning, and Reinforcement Learning.

Education Experience

PhD: City University of Hong Kong (2020.9 ~ 2024.10) Gruaduated from City University of Hong Kong (qs: 53) School: Data Science GPA: 4.05/4.3

Bachelor: Beihang University (2016.9 ~ 2020.7) Gruaduated from Beihang University (985, 211) School: Mathematics and Applied Mathematices GPA: 3.5/4.0

Reseach Interests

Large Language Models: Pre-training, Efficiency, E-commerce Knowledge, and Downstream Application.

Recommender Systems: Sequential Recommendation, Efficiency, Linear Complexity, and Generative Recommendation.

Federated Learning: Convergence Theory, and Federated Recommendation.

Reinforcement Learning: Hallucination Optimization.

Publications

• Expert Divergence Learning for MoE-based Language Models. ICLR, 2026 (CCF-A Conference, Accepted). https://arxiv.org/abs/2603.00054

• NEZHA: A Zero-sacrifice and Hyperspeed Decoding Architecture for Generative Recommendations. WWW, 2026 (CCF-A Conference, Accepted). https://arxiv.org/abs/2511.18793

• Unlocking Scaling Law in Industrial Recommendation Systems with a Three-step Paradigm based Large User Model. WSDM, 2026 (CCF-B Conference, Accepted). https://arxiv.org/abs/2502.08309

• How to inject knowledge efficiently? Knowledge Infusion Scaling Law for Pre-training Large Language Models. EMNLP, 2025 (CCF-B Conference, Accepted). https://arxiv.org/abs/2509.19371

• Benchmarking Large Language Models in E-commerce Leveraging Knowledge Graph. CIKM, 2025 (CCF-B Conference, Accepted). https://arxiv.org/abs/2503.15990

• ChineseEcomQA: A Scalable E-commerce Concept Evaluation Benchmark for Large Language Models. KDD, 2025 (CCF-A Conference, Accepted). https://arxiv.org/abs/2502.20196

• UQABench: Evaluating User Embedding for Prompting LLMs in Personalized Question Answering. KDD, 2025 (CCF-A Conference, Accepted). https://arxiv.org/abs/2502.19178

• Multi-task Offline Reinforcement Learning for Online Advertising in Recommender Systems. KDD, 2025 (CCF-A Conference, Accepted). https://arxiv.org/abs/2506.23090

• Efficient and Robust Regularized Federated Recommendation. CIKM, 2024 (CCF-B Conference, Accepted). https://arxiv.org/abs/2411.01540

• Analysis of Regularized Federated Learning. Neurocomputing (SCI-Q2 Journal, Accepted). https://arxiv.org/abs/2411.01548

• Deep Learning for Social Recommendation: A Survey. TOIS, 2023 (CCF-A Journal, Under Review).

• LinRec: Linear Attention Mechanism for Long-term Sequential Recommender Systems. SIGIR, 2023 (CCF-A Conference, Accepted). https://arxiv.org/abs/2411.01537