Weiwen Liu

Weiwen Liu

Tenure-track Associate Professor

Posts by Collection

publications

talks

Neural Re-ranking for Multi-stage Recommender Systems

Published:

Re-ranking is one of the most critical stages for multi-stage recommender systems (MRS), which re-orders the input ranking lists by modeling the cross-item interaction. Recent re-ranking methods have evolved into deep neural architectures due to the significant advances in deep learning. Neural re-ranking, therefore, has become a trending topic and many of the improved algorithms have demonstrated their use in industrial applications, enjoying great commercial success. The purpose of this tutorial is to explore some of the recent work on neural re-ranking, integrating them into a broader picture and paving ways for more comprehensive solutions for future research. In particular, we provide a taxonomy of current methods according to the objectives and training signals. We examine and compare these methods qualitatively and quantitatively, and identify some open challenges and future prospects.More information here

How Recommender Systems Can Benefit from Large Language Models: An Application Perspective

Published:

Recommender systems (RS) play a crucial role in mitigating information overload and delivering personalized content that caters to users’ diverse needs. The emergence of large language models (LLMs) offers a new horizon for redefining recommender systems with vast general knowledge and reasoning capabilities. In this talk, we trace the technical evolution of recommender systems, highlighting key advancements from deep learning to LLMs and LLM agents.

User Behavior Modeling with Deep Learning for Recommendation: Recent Advances

Published:

User Behavior Modeling (UBM) plays a critical role in user interest learning, and has been extensively used in recommender systems. The exploration of key interactive patterns between users and items has yielded significant improvements and great commercial success across a variety of recommendation tasks. This tutorial aims to offer an in-depth exploration of this evolving research topic. We start by reviewing the research background of UBM, paving the way to a clearer understanding of the opportunities and challenges. Then, we present a systematic categorization of existing UBM research works, which can be categorized into four different directions including Conventional UBM, Long-Sequence UBM, Multi-Type UBM, and UBM with Side Information. To provide an expansive understanding, we delve into each category, discussing representative models while highlighting their respective strengths and weaknesses. Furthermore, we elucidate on the industrial applications of UBM methods, aiming to provide insights into the practical value of existing UBM solutions. Finally, we identify some open challenges and future prospects in UBM. This comprehensive tutorial serves to provide a solid foundation for anyone looking to understand and implement UBM in their research or business. More information here

Making LLM Agents Work: Key Challenges and Solutions

Published:

Making LLM Agents Work: Addressing Key Bottlenecks:

  • ​Why is it hard for AI agents to do things correctly (not just talk)?
  • ​What makes it difficult for agents to use tools effectively, and how can we address that?
  • ​If agents get better at using tools, what additional capabilities can they unlock, and what major challenges will remain?

Towards Tool-augmented LLM Agents

Published:

With the rapid advancement of Artificial General Intelligence (AGI) research, large language models (LLMs) are gradually evolving from passive language generators into proactive problem solvers and task executors. In this process, the ability of LLMs to utilize external tools is widely regarded as a key bridge connecting language intelligence with real-world interaction. However, current LLM agents still face significant challenges in complex task scenarios, including limited generalization ability and low end-to-end execution accuracy in multi-turn interactions. This talk explores the challenges faced by mainstream LLM agent models in real-world applications and focuses on systematic approaches to building tool-use capabilities in LLMs. These include key technical pathways such as data synthesis, post-training, and iterative self-evolution. Additionally, the report presents cutting-edge progress in personalized tool use, encompassing methods such as user preference modeling, query completion, and proactive tool use, aiming to significantly reduce user interaction costs and improve the practicality and intelligence of LLM agents.

teaching

CS7346 Algorithms for Big Data

Graduate Course, Shanghai Jiao Tong University, 2026

This course introduces data mining and machine learning algorithms for analyzing big data, with a focus on various types of algorithms that can handle large-scale data. The course consists of seven chapters: Introduction to MapReduce and Spark, Introduction to Locally Sensitive Hashing Algorithms, Large Scale Machine Learning Algorithms, Data and Matrix Dimensionality Reduction, Graph Algorithms and Community Discovery, Introduction to Recommender Systems, and Temporal and Spatio-Temporal Data Mining. Big data algorithms and analytics refer to the methods or tools used to collect, process, and analyze data from different large data sets. These datasets may come from a variety of sources such as the Web, mobile applications, email, social media, and connected smart devices. They typically represent data generated at high speed in a variety of forms, ranging from structured (database tables, Excel sheets) to semi-structured (XML files, web pages) to unstructured (images, audio files). Therefore, this course covers most of the data types likely to be encountered in Big Data processing tasks in seven chapters, helping students understand how to use algorithms wisely to solve large-scale data problems.

CS7309 Reinforcement Learning Theory and Applications

Graduate Course, Shanghai Jiao Tong University, 2026

With the rapid advancement of artificial intelligence, machine learning is evolving beyond traditional predictive tasks to encompass more complex decision-making challenges, such as game AI, autonomous driving, and intelligent assistants. Unlike predictive tasks, decision-making tasks require agents not only to understand the current state but also to devise long-term strategies based on feedback from the environment. This demands stronger generalization and adaptive capabilities. Reinforcement Learning, as the core technology driving these capabilities, has attracted significant attention in recent years. In the era of AI powered by large language models, RL is recognized as a critical approach for enhancing model generalization and interaction performance. It holds great promise for the development of more intelligent, flexible AI systems and the advancement toward artificial general intelligence. This course offers a systematic introduction to the fundamental principles and key methodologies of Reinforcement Learning. Covered topics include Markov Decision Processes, Dynamic Programming, Temporal Difference Learning, Policy Gradient Methods, Deep Reinforcement Learning, Imitation Learning, and Multi-Agent Reinforcement Learning. The course emphasizes both theoretical understanding and practical implementation, encouraging students to build and optimize agents through programming to solve real-world problems. By completing this course, students will establish a strong foundation for further exploration at the forefront of artificial intelligence research.