Spatial-Temporal Trajectory Dataset

A curated collection of 19 spatial-temporal trajectory and mobility datasets for research in trajectory modeling, location prediction, and spatio-temporal point processes.

Dataset Overview

#	Dataset	Type	Records	Format	Task
1	COVID19	Spatio-temporal events	1,450 seq	Pickle	Point process modeling
2	Citibike	Bike trips	2,440 seq	Pickle	Mobility prediction
3	Crime	Categorical events	3,000 seq	Pickle	Marked point process
4	Earthquake	Seismic events	950 seq	Pickle	Spatio-temporal prediction
5	Independent	Point process	5,000 seq	Pickle	General STPP
6	Geolife Trajectories	GPS trajectories	18,670 traj	PLT text	Trajectory classification
7	Brightkite	LBSN check-ins	4.7M	TXT	Location prediction
8	Foursquare	LBSN check-ins + POI	33.3M	TXT	Venue recommendation
9	Gowalla	LBSN check-ins	6.4M	TXT	Mobility modeling
10	AgentMove	Multi-city check-ins	3.1M	CSV/JSON/NPZ	Cross-city transfer
11	Chengdu	Taxi trajectories	181K	CSV	Travel time estimation
12	Porto	Taxi trajectories	1.7M	CSV	Trajectory similarity
13	LIMP	Mobility + intent	960K	CSV	Intent-aware prediction
14	METR-LA	Traffic speed	207 sensors	HDF5	Traffic forecasting
15	PEMS08	Traffic flow	170 sensors	NPZ	Traffic forecasting
16	PEMS-Bay	Traffic speed	325 sensors	HDF5	Traffic forecasting
17	Tencent	Road network + GPS	320K	TXT	Map matching
18	Source	Transfer source domain	33.6M	CSV/TXT	Cross-city pretraining
19	Target	Transfer target domain	53.2M	CSV	Domain adaptation

Data Categories

Spatio-Temporal Point Processes

COVID19 · Citibike · Crime · Earthquake · Independent

LBSN Check-ins

Brightkite · Foursquare · Gowalla · AgentMove

Trajectories

Geolife Trajectories · Chengdu · Porto

Traffic

METR-LA · PEMS08 · PEMS-Bay · Tencent

Mobility & Transfer Learning

LIMP · Source · Target

Quick Start

Each dataset directory contains:

• README.md — detailed format description
• example/ — sample data files showing the exact format

License

Datasets are provided for research purposes. Refer to original sources for specific licenses.

Citation

If you use these datasets, please cite both the original dataset sources and this repository:

@misc{ai4datasynth2024,
  title     = {Spatial-Temporal Trajectory Dataset},
  author    = {AI4DataSynth},
  year      = {2024},
  publisher = {GitHub},
  url       = {https://github.com/AI4DataSynth/Spatial-Temporal-Trajectory-Dataset}
}