KDD 2024 | 时空数据(Spatio-temporal) Research论文总结

时空探索之旅

发布于 2024-11-19 16:46:00

1.3K0

文章被收录于专栏：时空探索之旅时空探索之旅

2024 KDD（ ACM SIGKDD Conference on Knowledge Discovery and Data Mining, 知识发现和数据挖掘会议）在2024年8月25日-29日在西班牙巴塞罗那举行。

本文总结了KDD2024有关时空数据(Spatial-temporal) 的相关论文，如有疏漏，欢迎大家补充。

时空数据Topic：时空（交通）预测,插补，气象预测，轨迹生成，预测，异常检测，信控优化，POI推荐等。

Research track中有3个session中与时空数据（城市计算）紧密相关：Urban data Ⅰ,Ⅱ 与 spatio-temporal data，还有一些其余session中有一些做的时空数据任务。

Urban data Ⅰ

Multi-Task Learning for Routing Problem with Cross-Problem Zero-Shot Generalization
CoSLight: Co-optimizing Collaborator Selection and Decision-making to Enhance Traffic Signal Control
CrossLight: Offline-to-Online Reinforcement Learning for Cross-City Traffic Signal Control
Online Preference Weight Estimation Algorithm with Vanishing Regret for Car-Hailing in Road Network
Rethinking Order Dispatching in Online Ride-Hailing Platforms
STONE: A Spatio-temporal OOD Learning Framework Kills Both Spatial and Temporal Shifts
UniST: A Prompt-empowered Universal Model for Urban Spatio-temporal Prediction

Urban Data II

ControlTraj: Controllable Trajectory Generation with Topology-Constrained Diffusion Model
Communication-efficient Multi-service Mobile Traffic Prediction by Leveraging Cross-service Correlations
Irregular Traffic Time Series Forecasting Based on Asynchronous Spatio-Temporal Graph Convolutional Networks
ITPNet: Towards Instantaneous Trajectory Prediction for Autonomous Driving
Profiling Urban Streets: A Semi-Supervised Prediction Model Based on Street View Imagery and Spatial Topology
Physics-informed Neural ODE for Post-disaster Mobility Recovery
DiffCrime: A Multimodal Conditional Diffusion Model for Crime Risk Map Inference

Spatio-temporal Data

Self-consistent Deep Geometric Learning for Heterogeneous Multi-source Spatial Point Data Prediction
MulSTE: A Multi-view Spatio-temporal Learning Framework with Heterogeneous Event Fusion for Demand-supply Prediction
Long-Term Vessel Trajectory Imputation with Physics-Guided Diffusion Probabilistic Model
RPMixer: Shaking Up Time Series Forecasting with Random Projections for Large Spatial-Temporal Data
Heterogeneity-Informed Meta-Parameter Learning for Spatiotemporal Time Series Forecasting
ImputeFormer: Low Rankness-Induced Transformers for Generalizable Spatiotemporal Imputation
Pre-Training Identification of Graph Winning Tickets in Adaptive Spatial-Temporal Graph Neural Networks

其它

DyPS: Dynamic Parameter Sharing in Multi-Agent Reinforcement Learning for Spatio-Temporal Resource Allocation
STEMO: Early Spatio-temporal Forecasting with Multi-Objective Reinforcement Learning
Multi-Scale Detection of Anomalous Spatio-Temporal Trajectories in Evolving Trajectory Datasets
ReFound: Crafting a Foundation Model for Urban Region Understanding upon Language and Visual Foundations
Where Have You Been? A Study of Privacy Risk for Point-of-Interest Recommendation
ROTAN: A Rotation-based Temporal Attention Network for Time-Specific Next POI Recommendation
Diffusion-Based Cloud-Edge-Device Collaborative Learning for Next POI Recommendations

Urban data Ⅰ

1. Multi-Task Learning for Routing Problem with Cross-Problem Zero-Shot Generalization

链接：https://arxiv.org/abs/2402.16891

ACM链接：https://dl.acm.org/doi/abs/10.1145/3637528.3672040

代码：https://github.com/FeiLiu36/MTNCO

作者：Fei Liu (City University of Hong Kong); Xi Lin (City University of Hong Kong); Zhenkun Wang (Southern University of Science and Technology); Qingfu Zhang (City University of Hong Kong); Tong Xialiang (Huawei Technologies Ltd.); Mingxuan Yuan (Huawei Technologies Ltd.)

关键词：供应链管理，零样本

2. CoSLight: Co-optimizing Collaborator Selection and Decision-making to Enhance Traffic Signal Control

链接：https://arxiv.org/abs/2405.17152

ACM链接：https://dl.acm.org/doi/abs/10.1145/3637528.3671998

代码：https://github.com/bonaldli/CoSLight

作者：Jingqing Ruan (Institute of Automation, Chinese Academy of Science, Chinese Academy of Sciences); Ziyue Li (EWI gGmbH, Sensetime Research); Hua Wei (Arizona State University); Haoyuan Jiang (Baidu Inc.); Jiaming Lu (Fudan University); Xuantang Xiong (Institute of Automation, Chinese Academy of Science, Chinese Academy of Sciences); Hangyu Mao (Sensetime Research); Rui Zhao (Sensetime Research, Qing Yuan Research Institute, Shanghai Jiao Tong University)

关键词：信控优化，多智能体强化学习

CoSLight

3. CrossLight: Offline-to-Online Reinforcement Learning for Cross-City Traffic Signal Control

链接：https://zhoujingbo.github.io/paper/2024CrossLightKDD.pdf

ACM链接：https://dl.acm.org/doi/abs/10.1145/3637528.3671927

作者：Qian Sun (Hong Kong University of Science and Technology); Rui Zha (University of Science and Technology of China); Le Zhang (Baidu Inc.); Jingbo Zhou (Baidu Inc.); Yu Mei (Baidu Inc.); Zhiling Li (Baidu Inc.); Hui Xiong (Hong Kong University of Science and Technology (Guangzhou), Hong Kong University of Science and Technology)

关键词：信控优化，offline2online

CrossLight

4. Online Preference Weight Estimation Algorithm with Vanishing Regret for Car-Hailing in Road Network

ACM链接：https://dl.acm.org/doi/abs/10.1145/3637528.3671664

作者：Yucen Gao (Shanghai Jiaotong University); Zhehao Zhu (Shanghai Jiaotong University); Mingqian Ma (Shanghai Jiaotong University); Fei Gao (Didi Global Inc.); Hui Gao (Didi Global Inc.); Yangguang Shi (Shandong University); Xiaofeng Gao (Shanghai Jiao Tong University)

关键词：偏好权重、状态在线学习、消失遗憾、动态确定性马尔可夫决策过程

PWC

5. Rethinking Order Dispatching in Online Ride-Hailing Platforms

ACM链接：https://dl.acm.org/doi/10.1145/3637528.3672028

作者：Zhaoxing Yang (Shanghai Jiao Tong University); Haiming Jin (Shanghai Jiao Tong University); Guiyun Fan (Shanghai Jiao Tong University); Min Lu (Didi Chuxing); Yiran Liu (Didi Chuxing); Xinlang Yue (Didi Chuxing); Hao Pan (Didi Chuxing); Zhe Xu (Didi Chuxing); Guobin Wu (Didi Chuxing); Qun Li (Didi Chuxing); Xiaotong Wang (Didi Chuxing); Jiecheng Guo (Didi Chuxing)

关键词：订单调度、网约车、多智能体强化学习

GRC

6. STONE: A Spatio-temporal OOD Learning Framework Kills Both Spatial and Temporal Shifts

ACM链接：https://dl.acm.org/doi/abs/10.1145/3637528.3671680

作者：Binwu Wang (University of Science and Technology of China); Jiaming Ma(University of Science and Technology of China); Pengkun Wang (Suzhou Institute for Advanced Research, University of Science and Technology of China); Xu Wang (Suzhou Institute for Advanced Research, University of Science and Technology of China); Yudong Zhang (University of Science and Technology of China); Zhengyang Zhou(Suzhou Institute for Advanced Research, University of Science and Technology of China); Yang Wang(University of Science and Technology of China)

关键词：时空OOD，分布偏移，因果图学习

STONE

7. UniST: A Prompt-empowered Universal Model for Urban Spatio-temporal Prediction

链接：https://arxiv.org/abs/2402.11838

ACM链接：https://dl.acm.org/doi/abs/10.1145/3637528.3671662

代码：https://github.com/tsinghua-fib-lab/UniST

作者：Yuan Yuan (Department of Electronic Engineering, BNRist, Tsinghua University); Jingtao Ding (Department of Electronic Engineering, BNRist, Tsinghua University); Jie Feng (Department of Electronic Engineering, BNRist, Tsinghua University); Depeng Jin (Department of Electronic Engineering, BNRist, Tsinghua University); Yong Li (Department of Electronic Engineering, BNRist, Tsinghua University)

关键词：时空（网格）预测，基础模型

论文精读 |2024[KDD]UniST: One-for-all城市时空预测模型，预训练+提示统一多种时空场景

UniSTUrban data Ⅱ

8. ControlTraj: Controllable Trajectory Generation with Topology-Constrained Diffusion Model

链接：https://arxiv.org/abs/2404.15380

ACM链接：https://dl.acm.org/doi/abs/10.1145/3637528.3671730

作者：Yuanshao Zhu (Southern University of Science and Technology, City University of Hong Kong); James Jianqiao Yu (University of York); Xiangyu Zhao (City University of Hong Kong); Qidong Liu (Xi’an Jiao Tong University, City University of Hong Kong); Yongchao Ye (City University of Hong Kong); Wei Chen (Hong Kong University of Science and Technology (Guangzhou)); Zijian Zhang (Jilin University, City University of Hong Kong); Xuetao Wei (Southern University of Science and Technology); Yuxuan Liang (The Hong Kong University of Science and Technology (Guangzhou))

关键词：轨迹生成，扩散模型

ControlTraj

9. Communication-efficient Multi-service Mobile Traffic Prediction by Leveraging Cross-service Correlations

ACM链接：https://dl.acm.org/doi/abs/10.1145/3637528.3671730

作者：Zhiying Feng (School of Computer Science and Engineering, Sun Yat-sen University); Qiong Wu (Department of Electronic and Computer Engineering, Hong Kong University of Science and Technology); Xu Chen (School of Computer Science and Engineering, Sun Yat-sen University)

关键词：交通预测，跨设备聚合，迁移学习，图卷积

CsASTGCN

10. Irregular Traffic Time Series Forecasting Based on Asynchronous Spatio-Temporal Graph Convolutional Networks

链接：https://arxiv.org/abs/2308.16818

ACM链接：https://dl.acm.org/doi/abs/10.1145/3637528.3671665

作者：Weijia Zhang (HKUST(GZ)); Le Zhang(Baidu Research); Jindong Han (HKUST); Hao Liu (HKUST(GZ), HKUST); Yanjie Fu (Arizona State University); Jingbo Zhou (Baidu Research); Yu Mei(Baidu Inc.); Hui Xiong (HKUST(GZ), HKUST)

关键词：不规则时序交通预测，时空图神经网络

ASeer

11. ITPNet: Towards Instantaneous Trajectory Prediction for Autonomous Driving

链接：https://openreview.net/forum?id=mDIXfHvoqH

ACM链接：https://dl.acm.org/doi/abs/10.1145/3637528.3671681

作者：Rongqing Li (Beijing Institute of Technology); Changsheng Li (Beijing Institute of Technology); Yuhang Li (Beijing Institute of Technology); Hanjie Li (Beijing Institute of Technology); Yi Chen (Beijing Institute of Technology); Ye Yuan (Beijing Institute of Technology); Guoren Wang (Beijing Institute of Technology)

关键词：轨迹预测，自动驾驶

ITPNet

12. Profiling Urban Streets: A Semi-Supervised Prediction Model Based on Street View Imagery and Spatial Topology

ACM链接：https://dl.acm.org/doi/abs/10.1145/3637528.3671918

作者：Meng Chen (School of Software, Shandong University); Zechen Li (School of Software, Shandong University); Weiming Huang (School of Computer Science and Engineering, Nanyang Technological University); Yongshun Gong (School of Software, Shandong University); Yilong Yin (School of Software, Shandong University)

关键词：城市街道剖析，街景图像，街道表征学习，大语言模型

USPM

13. Physics-informed Neural ODE for Post-disaster Mobility Recovery

ACM链接：https://dl.acm.org/doi/abs/10.1145/3637528.3672027

作者：Jiahao Li (Shenzhen International Graduate School, Tsinghua University); Huandong Wang (Department of Electronic Engineering, Tsinghua University); Xinlei Chen (Shenzhen International Graduate School, Tsinghua University, Pengcheng Laboratory)

关键词：移动恢复，物理驱动，神经常微分方程

14. DiffCrime: A Multimodal Conditional Diffusion Model for Crime Risk Map Inference

ACM链接：https://dl.acm.org/doi/abs/10.1145/3637528.3671843

作者：Shuliang Wang (Beijing Institute of Technology); Xinyu Pan (Beijing Institute of Technology); Sijie Ruan (Beijing Institute of Technology); Haoyu Han (Beijing Institute of Technology); Ziyu Wang (Beijing Institute of Technology); Hanning Yuan (Beijing Institute of Technology); Jiabao Zhu (Beijing Institute of Technology); Qi Li (Beijing Institute of Technology)

关键词：犯罪风险图推理，多模态，扩散模型

DiffCrimeSpatio-temporal Data

15. Self-consistent Deep Geometric Learning for Heterogeneous Multi-source Spatial Point Data Prediction

链接：https://arxiv.org/abs/2407.00748

ACM链接：https://dl.acm.org/doi/abs/10.1145/3637528.3671737

作者：Dazhou Yu (Emory University); Xiaoyun Gong (Emory University); Yun Li (Emory University); Meikang Qiu (Augusta University); Liang Zhao (Emory University)

关键词：空间点数据预测，多源空间数据，几何深度学习

KDD24 通用的多源空间点数据预测：以PM2.5预测为例

DMSP

16. MulSTE: A Multi-view Spatio-temporal Learning Framework with Heterogeneous Event Fusion for Demand-supply Prediction

ACM链接：https://dl.acm.org/doi/abs/10.1145/3637528.3672030

作者：Li Lin (Southeast University); Zhiqiang Lu (Southeast University); Shuai Wang (Southeast University); Yunhuai Liu (Peking University); Zhiqing Hong (Rutgers University); Haotian Wang (JD Logistics); Shuai Wang (Southeast University)

关键词：供需预测，事件表示，图神经网络，时空图

MulSTE

17. Long-Term Vessel Trajectory Imputation with Physics-Guided Diffusion Probabilistic Model

ACM链接：https://dl.acm.org/doi/abs/10.1145/3637528.3672086

作者：Zhiwen Zhang (School of Artificial Intelligence, Jilin University); Zipei Fan (School of Artificial Intelligence, Jilin University); Zewu Lv (School of Artificial Intelligence, Jilin University); Xuan Song (School of Artificial Intelligence, Jilin University, Research Institute of Trustworthy Autonomous Systems, Southern University of Science and Technology (SUSTech)); Ryosuke Shibasaki (Research & Development Department, LocationMind Inc.)

关键词：船舶轨迹插补，扩散模型、轨迹嵌入、物理引导

18. RPMixer: Shaking Up Time Series Forecasting with Random Projections for Large Spatial-Temporal Data

链接：https://www.researchgate.net/publication/381321813_RPMixer_Shaking_Up_Time_Series_Forecasting_with_Random_Projections_for_Large_Spatial-Temporal_Data

ACM链接：https://dl.acm.org/doi/abs/10.1145/3637528.3671881

作者：Chin-Chia Michael Yeh (Visa Research); Yujie Fan (Visa Research); Xin Dai (Visa Research); Uday Singh Saini (Visa Research); Vivian Lai (Visa Research); Prince Aboagye (Visa Research); Junpeng Wang (Visa Research); Huiyuan Chen (Visa Research); Yan Zheng (Visa Research); Zhongfang Zhuang (Visa Research); Liang Wang (Visa Research); Wei Zhang (Visa Research)

关键词：时空预测，大规模时空图

RPMixer

19. Heterogeneity-Informed Meta-Parameter Learning for Spatiotemporal Time Series Forecasting

链接：https://arxiv.org/abs/2405.10800

ACM链接：https://dl.acm.org/doi/abs/10.1145/3637528.3671961

代码：https://github.com/XDZhelheim/HimNet

作者：Zheng Dong (Southern University of Science and Technology); Renhe Jiang (The University of Tokyo); Haotian Gao (The University of Tokyo); Hangchen Liu (Southern University of Science and Technology); Jinliang Deng (Hong Kong University of Science and Technology); Qingsong Wen (Squirrel AI); Xuan Song (Jilin University, Southern University of Science and Technology)

关键词：时空预测，异质性，元参数学习

KDD2024 | 解决时空建模异质问题——基于时空表征的个性化参数时空网络

HimNet

20. ImputeFormer: Low Rankness-Induced Transformers for Generalizable Spatiotemporal Imputation

链接：https://arxiv.org/abs/2312.01728

ACM链接：https://dl.acm.org/doi/abs/10.1145/3637528.3671751

代码：https://github.com/tongnie/ImputeFormer

作者：Tong Nie (Tongji University); Guoyang Qin (Tongji University); Wei Ma (The Hong Kong Polytechnic University); Yuewen Mei (Tongji University); Jian Sun (Tongji University)

关键词：时空插补，低秩建模

论文精读|2024[KDD]ImputeFormer: 用于广义时空补全的低秩诱导的Transformer

ImputeFormer

21. Pre-Training Identification of Graph Winning Tickets in Adaptive Spatial-Temporal Graph Neural Networks

链接：https://arxiv.org/abs/2406.08287

ACM链接：https://dl.acm.org/doi/abs/10.1145/3637528.3671912

代码：https://anonymous.4open.science/r/paper-1430

作者：Wenying Duan (Jiangxi Provincial Key Laboratory of Intelligent Systems and Human-Machine Interaction, Nanchang University); Tianxiang Fang (Nanchang University); Hong Rao (School of Software, Nanchang University); Xiaoxi He (Faculty of Science and Technology, University of Macau)

关键词：时空图神经网络，彩票假说

其他

22. DyPS: Dynamic Parameter Sharing in Multi-Agent Reinforcement Learning for Spatio-Temporal Resource Allocation

ACM链接：https://dl.acm.org/doi/abs/10.1145/3637528.3672052

作者：Jingwei Wang (Department of Electronic Engineering, Beijing National Research Center for Information Science and Technology, Tsinghua University); Qianyue Hao (Department of Electronic Engineering, Beijing National Research Center for Information Science and Technology, Tsinghua University); Wenzhen Huang (Department of Electronic Engineering, Beijing National Research Center for Information Science and Technology, Tsinghua University); Xiaochen Fan (Department of Electronic Engineering, Beijing National Research Center for Information Science and Technology, Tsinghua University); Zhentao Tang (Huawei Noah’s Ark Lab); Bin Wang (Huawei Noah’s Ark Lab); Jianye Hao (Huawei Noah’s Ark Lab, Tianjin University); Yong Li (Department of Electronic Engineering, Beijing National Research Center for Information Science and Technology, Tsinghua University)

关键词：多智能体强化学习，时空资源分配

DyPS

23. STEMO: Early Spatio-temporal Forecasting with Multi-Objective Reinforcement Learning

链接：https://arxiv.org/abs/2406.04035

ACM链接：https://dl.acm.org/doi/abs/10.1145/3637528.3671922

代码：https://github.com/coco0106/MO-STEP

作者：Wei Shao (Data61, CSIRO); Yufan Kang (RMIT University); Ziyan Peng (Xidian University); Xiao Xiao (Xidian University); Lei Wang (Zhejiang University); Yuhui Yang (Xidian University); Flora D. Salim (University of New South Wales)

关键词：早期预测，强化学习

论文精读 | 2024[KDD]基于多目标强化学习的时空早期预测

STEMO

24. Multi-Scale Detection of Anomalous Spatio-Temporal Trajectories in Evolving Trajectory Datasets

ACM链接：https://dl.acm.org/doi/abs/10.1145/3637528.3671874

作者：Chenhao Wang (University of Electronic Science and Technology of China); Lisi Chen (University of Electronic Science and Technology of China); Shuo Shang (University of Electronic Science and Technology of China); Christian S. Jensen (Aalborg University); Panos Kalnis (King Abdullah University of Science and Technology)

关键词：轨迹异常检测

MST-OATD

25. ReFound: Crafting a Foundation Model for Urban Region Understanding upon Language and Visual Foundations

链接：https://zhoujingbo.github.io/paper/2024ReFoundKDD.pdf

ACM链接：https://dl.acm.org/doi/abs/10.1145/3637528.3671992

作者：Congxi Xiao (School of Computer Science and Technology, University of Science and Technology of China); Jingbo Zhou (Business Intelligence Lab, Baidu Research); Yixiong Xiao (Business Intelligence Lab, Baidu Research); Jizhou Huang (Baidu Inc.); Hui Xiong (Thrust of Artificial Intelligence, The Hong Kong University of Science and Technology (Guangzhou), Department of Computer Science and Engineering, The Hong Kong University of Science and Technology)

关键词：城市区域理解，多模态

ReFound

26. Where Have You Been? A Study of Privacy Risk for Point-of-Interest Recommendation

ACM链接：https://dl.acm.org/doi/abs/10.1145/3637528.3671758

代码：https://github.com/KunlinChoi/POIPrivacy

作者：Kunlin Cai (University of California, Los Angeles); Jinghuai Zhang (University of California, Los Angeles); Zhiqing Hong (Rutgers University); William Shand (University of California, Los Angeles); Guang Wang (Florida State University); Desheng Zhang (Rutgers University); Jianfeng Chi (Meta); Yuan Tian (University of California, Los Angeles)

关键词：POI推荐，隐私保护，关系推理

27.ROTAN: A Rotation-based Temporal Attention Network for Time-Specific Next POI Recommendation

ACM链接：https://dl.acm.org/doi/abs/10.1145/3637528.36718098

代码：https://github.com/ruiwenfan/ROTAN

作者：Shanshan Feng (Centre for Frontier AI Research, ASTAR, Institute of High Performance Computing, ASTAR); Feiyu Meng (University of Electronic Science and Technology of China); Lisi Chen (University of Electronic Science and Technology of China); Shuo Shang (University of Electronic Science and Technology of China); Yew Soon Ong (Centre for Frontier AI Research, A*STAR, Nanyang Technological University)

关键词：POI推荐，旋转（Rotations）

ROTAN

28. Diffusion-Based Cloud-Edge-Device Collaborative Learning for Next POI Recommendations

链接：https://arxiv.org/abs/2405.13811

ACM链接：https://dl.acm.org/doi/abs/10.1145/3637528.3671743

代码：https://github.com/ruiwenfan/ROTAN

作者：Jing Long (The University of Queensland); Guanhua Ye (Beijing University of Posts and Telecommunications); Tong Chen (The University of Queensland); Yang Wang (Hefei University of Technology); Meng Wang (Hefei University of Technology); Hongzhi Yin (The University of Queensland)

关键词：POI推荐，on-device，扩散模型

DCPR