SIGMOD 2026刚刚在印度班加罗尔落下帷幕，5月31日至6月5日的议程里，时空数据（Spatial Temporal）方向涌现了不少值得关注的工作。下面就来梳理一下这一领域的5篇论文，涵盖轨迹压缩、轨迹相似度计算、地理空间数据库、多模态大模型、数据清洗等热点方向。如有遗漏，欢迎补充。
1. DimWea ver: Dimensional Wea ved Trajectory Compression |
|---|
1 DimWea ver: Dimensional Wea ved Trajectory Compression
链接：https://dl.acm.org/doi/10.1145/3802040
代码：https://github.com/xkitsios/DimWea ver
作者：Xenophon Kitsios (Athens University of Economics and Business)*; Panagiotis Liakos (Athens University of Economics and Business); Katia Papakonstantinopoulou (Athens University of Economics and Business); Nikos Giatrakos (Technical University of Crete); Yannis Kotidis (Athens University of Economics and Business)
关键词：轨迹压缩
2 GoodTP: An Effective Data Selection Framework for Enhancing Trajectory Similarity Learning via Monte Carlo Tree Search
链接：https://dl.acm.org/doi/10.1145/3802067
代码：https://github.com/yuanhaitao/GoodTP/t
作者：Haitao Yuan (Nanyang Technological University)*; Gao Cong (Nanyang Technological University)
关键词：数据选择，轨迹相似度计算
3 GeoKGM: A Multimodal Large Language Model for Zero-Shot Knowledge Graph Completion in Geospatial Databases
链接：https://dl.acm.org/doi/abs/10.1145/3769796
作者：Zhihan Zheng (Beijing University of Posts and Telecommunications)*; Haitao Yuan (Nanyang Technological University); Minxiao Chen (Beijing University of Posts and Telecommunications); Nan Jiang (Beijing University of Posts and Telecommunications); Haoning Wang (National University of Singapore); Shangguang Wang (Beijing University of Posts and Telecommunications)
关键词：多模态大模型，零样本知识图谱补全，地理空间数据库
4 3dSAGER: Geospatial Entity Resolution over 3D Objects
链接：https://dl.acm.org/doi/10.1145/3769751
代码：https://github.com/BarGenossar/3dSAGER
作者：Bar Genossar (Technion -- Israel Institute of Technology)*; Sagi Dalyot (Technion -- Israel Institute of Technology); Roee Shraga (Worcester Polytechnic Institute); A vigdor Gal (Technion -- Israel Institute of Technology)
关键词：地理空间实体解析，实体匹配，块划分
5 From Suspicious Errors to Valid Data: On Repairing Spatio-Temporal Data via Spatial and Temporal Dependencies
链接：https://dl.acm.org/doi/abs/10.1145/3769794
代码：https://github.com/cookiesvivi/STRhub
作者：Weiwei Deng (Nankai University); Yu Sun (Nankai University)*; Shaoxu Song (Tsinghua University); Xiaojie Yuan (Nankai University)
关键词：时空数据，数据修复，数据清洗
