Personal Update:
"Keep a peaceful heart"

Research Philosophy: Scalable and Principled - The Bitter Lesson

About Me

Hi! I am Wei Chen, a Ph.D. Student (Since 2023.09) at Hong Kong University of Science and Technology, Guangzhou. I am fortunate to be mentored by Prof. Yuxuan Liang and Prof. Xiaofang Zhou.

My past research journey includes experiences at Microsoft Research Asia (mentored by Justin Ding), iFLYTEK Research Institute (advised by Dr. Weitai Zhang), and Ocean University of China (mentored by Prof. Yanwei Yu).

My research interests revolve around modern AI technology and its wide range of applications, with the core being more efficient, green, and friendly AI for humans, cities, and science (recently, I have been particularly interested in opportunities to combine LLM with all aspects). Specifically, they include:

Data-centric: (more efficient): NIPS'24
Algorithm-centric: (more durable): ICML'24, IJCAI'24
Models-centric: (more transparent): NIPS'24, TOIS'24, TKDD'23, IJCAI'22
Applications: AI for social good: WWW'24, KDD'24

To contribute more ideas to the research community, I am trying to make the datasets and codes of our research publicly available on GitHub.

Selected News

(For the full list of news please check the News tab.)

I was selected as a "Top Reviewer" for NIPS'24 and was rewarded with a complimentary registration.
Sep. 2024: Two papers about Multimodal Spatio-temporal Datasets (Earth Science) / Time Series Forecasting Models (Chaos Theory) was accepted to NIPS'24.
May. 2024: One paper about "ControlTraj: Controllable Trajectory Generation with Topology-Constrained Diffusion Model" was accepted to KDD'24.
May. 2024: One paper about "Position paper: What Can Large Language Models Tell Us about Time Series Analysis" was accepted to ICML'24.
Apr. 2024: One paper about "Towards Robust Trajectory Representations: Isolating Environmental Confounders with Causal Learning" was accepted to IJCAI'24.
Mar. 2024: Our paper about "Deep Learning for Trajectory Data Management and Mining: A Survey and Beyond" was reported by Zhuan Zhi News.
Mar. 2024: We systematically review the progress of deep learning for trajectory computation (trajectory data management and mining) over the past decade. Please check out our latest Survey and Website.
Jan. 2024: One paper about "UrbanCLIP: Learning Text-enhanced Urban Region Profiling with Contrastive Language-Image Pretraining from the Web" was accepted to WWW'24.
Jan. 2024: One paper about "MCN4Rec: Multi-Level Collaborative Neural Network for Next Location Recommendation" was accepted to ACM Transactions on Information Systems (TOIS).
Nov. 2023: One paper about "Trajectory-User Linking via Hierarchical Spatio-Temporal Attention Networks" was accepted to ACM Transactions on Knowledge Discovery from Data (TKDD).
Sep. 2023: Starting my Ph.D. journey at the HKUST(GZ)!

vistors & views since September 2023