Published in The 25th ACM SIGKDD International Conference on Knowledge Discovery and Data Mining (SIGKDD 2019). (Anchorage, AK), 2019
This paper is a new framework to discovery short and long-term patterns over geo-spatiotemporal data (e.g. traffic, weather, etc.). Short-term patterns show propagation of geo-spatiotemporal entities which cause other entities to happen (e.g. rain –> accident –> traffic jam). Long-term patterns show influence of a temporally long-term entitiy on its spatiotemporal neighbhorhood (e.g. major construction –> more traffic issue).