STGEN: Deep Continuous-Time Spatiotemporal Graph Generation

Chen Ling, Hengning Cao, Liang Zhao

Abstract

Spatiotemporal graph generation has realistic social significance since it unscrambles the underlying distribution of spatiotemporal graphs from another perspective and fuels substantial spatiotemporal data mining tasks. Generative models for temporal and spatial networks respectively cannot be easily generalized to spatiotemporal graph generation due to their incapability of capturing: 1) mutually influenced graph and spatiotemporal distribution, 2) spatiotemporal-validity constraints, and 3) characteristics of multi-modal spatiotemporal properties. To this end, we propose a generic and end-to-end framework for spatiotemporal graph generation (STGEN) that jointly captures the graph, temporal, and spatial distributions of spatiotemporal graphs. Particularly, STGEN learns the multi-modal distribution of spatiotemporal graphs via learning the distribution of spatiotemporal walks based on a new heterogeneous probabilistic sequential model. Auxiliary activation layers are proposed to retain the spatiotemporal validity of the generated graphs. In addition, a new boosted strategy for the ensemble of discriminators is proposed to distinguish the generated and real spatiotemporal walks from multi-dimensions and capture the combinatorial patterns among them. Finally, extensive experiments are conducted on both synthetic/real-world spatiotemporal graphs and demonstrated the efficacy of the proposed model.
BibTeX
@inproceedings{ling2022stgen,
  author = {Chen Ling and Hengning Cao and Liang Zhao},
  title = {{STGEN: Deep Continuous-Time Spatiotemporal Graph Generation}},
  booktitle = {ECML/PKDD},
  year = {2022},
  doi = {10.1007/978-3-031-26409-2\_21},
}

Loading related content...

Figure